- Email: [email protected]
Different strokes for different folks: Effects of regulatory mode complementarity and task complexity on performance
Personality and Individual Differences 89 (2016) 134–142
Contents lists available at ScienceDirect
Personality and Individual Differences journal homepage: www.elsevier.com/locate/paid
Different strokes for different folks: Effects of regulatory mode complementarity and task complexity on performance☆ Marina Chernikova a,⁎, Calogero Lo Destro b, Romina Mauro b, Antonio Pierro b, Arie W. Kruglanski a, E. Tory Higgins c a b c
University of Maryland, United States Universita di Roma “La Sapienza”, Italy Columbia University, United States
a r t i c l e
i n f o
Article history: Received 21 July 2015 Received in revised form 1 October 2015 Accepted 3 October 2015 Available online 22 October 2015 Keywords: Self-regulation Locomotion Assessment Regulatory mode Performance
a b s t r a c t We examine how task features interact with individuals' regulatory modes in determining performance. Effective goal pursuit normally occurs when high locomotion (the regulatory mode concerned with motion from state to state) and high assessment (the regulatory mode concerned with critical evaluation) work together (Kruglanski et al., 2000). However, there may be situations in which this high–high combination is unnecessary or even detrimental to good performance. We hypothesized that on simple tasks, high locomotion and low assessment should lead to the best performance; on complex tasks, however, high locomotion and high assessment should lead to the best performance. We tested these hypotheses in two empirical studies, one carried out in an organizational setting, the other in the lab. In the first study, we measured individuals' locomotion and assessment tendencies, asked them to rate the complexity of their daily work tasks, and obtained measures of their job performance from their supervisors. In the second study, we measured individuals' locomotion and assessment tendencies, manipulated the task complexity of an inbox task they had to complete, and measured their performance on that task. Both studies provided support for our hypotheses. These results offer important insights regarding the effects of regulatory mode on performance. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Consider a brief tale of two friends. One of them, Michael, is always energetic and quick at his work. The other friend, Max, is also energetic and quick; in addition to this, Max also likes to methodically analyze his options and decide upon the best course of action before he starts. When they are given a task to complete, Michael begins right away and speeds through the task; Max, on the other hand, spends some time thinking deeply about the task and how to best complete it, then also gets going and completes the task quickly. It seems fairly straightforward to assume that Max (who focuses on both careful thought and swift action) will generally outperform Michael (who focuses only on swift action). But is this intuition correct? In order to answer this question, it is essential to distinguish between two basic aspects of self-regulation: locomotion and assessment. Regulatory mode theory postulates that every act of self-regulation involves two components: evaluating which goal or means to select during goal pursuit, and committing energy to engaging in the chosen ☆ We would like to thank Katarzyna Jasko and Maxim Milyavsky for their valuable comments on an earlier draft of the paper. ⁎ Corresponding author. E-mail address: [email protected] (M. Chernikova).
http://dx.doi.org/10.1016/j.paid.2015.10.011 0191-8869/© 2015 Elsevier Ltd. All rights reserved.
option (Higgins, Kruglanski, & Pierro, 2003; Kruglanski et al., 2000). These motivational components, or regulatory modes, are referred to as assessment and locomotion: assessment is concerned with comparison and evaluation, and locomotion is concerned with making progress toward a goal. The two regulatory modes can be measured as individual traits (Kruglanski et al., 2000) or manipulated as state variables (Avnet & Higgins, 2003). Furthermore, locomotion and assessment are assumed to be functionally independent, so that an individual can be high on both, low on both, or high on one and low on the other (Higgins et al., 2003; Kruglanski et al., 2000). An individual's regulatory mode can influence how she carries out a wide variety of activities in her daily life, ranging from the types of goals she selects (e.g.Mannetti, Pierro, Higgins, & Kruglanski, 2012; Orehek, Mauro, Kruglanski, & van der Bles, 2012) to the manner in which she pursues those goals (e.g. Avnet & Higgins, 2003; Orehek & Vazeou-Nieuwenhuis, 2013). The implications of each regulatory mode for goal pursuit will be elaborated in the following sections. Locomotion is defined as “the aspect of self-regulation concerned with movement from state to state, and with committing the psychological resources that will initiate and maintain goal-related movement in a straightforward and direct manner, without undue distractions or delays” (Kruglanski et al., 2000, p. 794). As such, the essential motivation of individuals high on locomotion is to experience psychological
M. Chernikova et al. / Personality and Individual Differences 89 (2016) 134–142
movement. High locomotors emphasize “hurrying,” “doing,” and “getting on with it,” since all of those allow them to avoid standing still and doing nothing (Higgins et al., 2003). Previous research suggests that the locomotion orientation can influence many aspects of goal pursuit. For example, high locomotors avoid procrastinating on goals and tasks: locomotion is negatively associated both with scores on a procrastination scale (which measures the tendency to delay task initiation or completion), and with insurance workers' actual procrastination as measured over a three month period (Pierro, Giacomantonio, Pica, Kruglanski, & Higgins, 2011). High locomotors tend to complete tasks as quickly as possible (even when such a focus on speed comes at the expense of accuracy; Kruglanski et al., 2000; Mauro, Pierro, Mannetti, Higgins, & Kruglanski, 2009). For instance, when engaged in a proofreading task, high (vs. low) locomotors take significantly less time to finish the task (Kruglanski et al., 2000). High locomotors exhibit a greater ability to stay focused on a task and avoid becoming distracted: locomotion is positively associated both with perseverance (as measured by a scale tapping the ability to sustain effort in the face of diversity) and resistance to temptation (as measured by participants' reports of how likely they would be to put off studying for an important exam; Pierro et al., 2011). Individuals who are high on locomotion are better at managing the time they devote to their goals: high locomotors have an increased proficiency at setting goals and priorities, a greater preference for organization, and greater perceived control over their time (Amato, Pierro, Chirumbolo, & Pica, 2014). High locomotors are more likely to select high expectancy (vs. high value) options during goal pursuit; in other words, locomotors tend to select means which serve only one goal (and thus may have a greater expectancy of attaining that goal) rather than means which serve multiple goals (and may have a lower expectancy of attaining any one of them; Orehek et al., 2012). Individuals who are high (vs. low) on locomotion are more likely to follow through on initial goals they set (such as attending fitness classes; Mannetti et al., 2012). Lastly, when selecting a means to their goal, high locomotors prefer to make simultaneous evaluations of all possible alternatives (rather than engaging in a sequential series of comparisons; Avnet & Higgins, 2003). Thus, locomotion can influence goal pursuit in a wide variety of ways. Assessment is defined as “the comparative aspect of self-regulation concerned with critically evaluating entities or states, such as goals or means, in relation to alternatives in order to judge relative quality” (Kruglanski et al., 2000, p. 794). The essential motivation of individuals high on assessment is to “make the right choice”; as such, high assessors are preoccupied with making comparisons and critical evaluations to ensure that they arrive at the correct decision before moving forward (Higgins et al., 2003). Like locomotion, assessment can also impact various aspects of individuals' goal pursuit. For instance, individuals who are high on assessment tend to procrastinate on goals and tasks: assessment is positively correlated both with scores on a procrastination scale, and with insurance workers' actual procrastination as measured over a three month period (Pierro et al., 2011). High assessors are slower to complete tasks, but are also more accurate at those tasks (Kruglanski et al., 2000; Mauro et al., 2009). For example, high (vs. low) assessors were found to be significantly more accurate when completing a proofreading task (Kruglanski et al., 2000). High assessors also tend to have greater concern over potential mistakes during goal pursuit (measured with items such as “I should be upset if I make a mistake”), higher personal standards for their task performance (measured with items such as “I expect higher performance in my daily tasks than most people”), and greater doubts about whether they are making the right choice (measured with items such as “I tend to get behind in my work because I repeat things over and over”; Pierro et al., 2011). High assessors are more likely to select high value (vs. high expectancy) options during goal pursuit: they tend to choose means which serve multiple goals (and can thus produce more overall value in terms of the amount of goals they attain) rather than means which serve only one goal (and
135
which therefore produce less overall value; Orehek et al., 2012). High assessors have a greater tendency to form behavioral goals (such as the goal to attend gym classes) based on perceived social norms (Mannetti et al., 2012). Lastly, high assessors prefer to select a means to their goal by making a sequential series of comparisons among the possible alternatives (rather than simultaneously evaluating all their options; Avnet & Higgins, 2003). In summary, previous research has shown that both assessment and locomotion regulatory mode are relevant to goal pursuit in a variety of domains. As described earlier, regulatory mode theory posits that locomotion and assessment are independent; therefore, if an individual is low on one of the orientations, he will not necessarily be high on the other (Higgins et al., 2003; Kruglanski et al., 2000). This independence assumption implies that different combinations of locomotion and assessment can have very different effects on goal pursuit and attainment. Prior research on the interactive effects of locomotion and assessment has generally focused on the beneficial effects of being high in both (Kruglanski, Pierro, Mannetti, & Higgins, 2013). In this vein, individuals who are high in both locomotion and assessment have been shown to have higher GPAs (Kruglanski et al., 2000), greater chances of successfully completing an elite army training unit (Kruglanski et al., 2000), and better work performance (Pierro, Pica, Mauro, Kruglanski & Higgins, 2012). However, no previous studies have examined whether being high on both locomotion and assessment is always beneficial, or whether there are situations in which being high on both can actually be detrimental to performance. The aim of this paper is to address this gap in the literature by focusing on one potential moderator of the effects of regulatory mode on performance: task complexity. Cognitive task complexity is defined as any attentional, memory, reasoning, or other information processing demands that are imposed by the structure of a task (Robinson, 2001). We therefore conceptualize task complexity as a function of objective task characteristics (March & Simon, 1958; Schwab & Cummings, 1976). Some characteristics of more (vs. less) complex cognitive tasks include a greater amount of information to process (Campbell, 1988; Robinson, 2001; Schroder, Driver, & Streufert, 1967), the presence of multiple ways to attain the goal (Terborg & Miller, 1978), uncertainty of outcomes (March & Simon, 1958), more task dimensions requiring attention (Schroder et al., 1967), and higher rate of information change (Schroder et al., 1967). Given that locomotors have been shown to exhibit a speed-accuracy tradeoff (in which high locomotors are faster, but high assessors are more accurate; Kruglanski et al., 2000), we suggest that individuals who are high only on locomotion, only on assessment, or on both locomotion and assessment may differ in their aptitude for completing tasks that differ on some of the aforementioned characteristics. When a cognitive task involves less information processing, only a single way to achieve the goal, more certain outcomes, fewer dimensions that require attention, or a lower rate of information change (in other words, when the task is simple), it requires less thought to execute successfully (Campbell, 1988). As such, a simple task can be accomplished merely by starting it as soon as possible and moving continually toward its attainment. Individuals high on locomotion prefer to avoid procrastinating (Pierro et al., 2011) and tend to move continuously toward their goal, disliking obstacles or interruptions (Kruglanski, Pierro, & Higgins, 2015; Kruglanski et al., 2000). Thus, high locomotors should be particularly well-suited to performing simple tasks. On the other hand, individuals high on assessment tend to gather information extensively and compare many alternatives before coming to a decision (Kruglanski et al., 2000). Since simple tasks do not require such information gathering for optimal performance, high assessors are not necessarily well-suited to performing such tasks. Tasks which include a greater amount of information to process, the presence of several ways to attain the goal, uncertain outcomes, many task dimensions that require attention, or a higher rate of information change (in other words, cognitively complex tasks) require extensive information gathering and processing (Campbell, 1988). As such,
136
M. Chernikova et al. / Personality and Individual Differences 89 (2016) 134–142
complex tasks are best accomplished by critically evaluating the task and selecting a proper strategy before proceeding. Individuals high on assessment are therefore particularly well-suited to performing complex tasks, as they enjoy critical thought and careful evaluation (Higgins et al., 2003). In addition to this, locomotion also plays a role in complex task performance. Performance on all tasks, including complex ones, is improved when one starts the task without procrastinating and when one moves toward the goal continuously—both of which are characteristics of high locomotors (Kruglanski et al., 2000; Kruglanski et al., 2015; Pierro et al., 2011). Thus, individuals who are high on both locomotion and assessment should be the most suited for performing complex tasks. In accordance with the logic outlined above, we propose the following hypotheses: 1. When a task is simple, the combination of high locomotion and low assessment should lead to the best task performance. 2. When a task is complex, the combination of high locomotion and high assessment should lead to the best task performance. We tested these hypotheses in two empirical studies, one carried out in an organizational setting, the other in the lab. In the first study, we measured individuals' chronic locomotion and assessment tendencies, asked them to rate the complexity of their daily tasks at work, and obtained measures of their job performance from their supervisors. In the second study, we measured individuals' chronic locomotion and assessment tendencies, manipulated the task complexity of an inbox task they had to complete in the lab, and measured how well they performed on that task. In these two studies we focus specifically on information load, or the amount of information to be processed, as an operationalization of task complexity. However, we argue that the logic outlined here should apply to all types of complex tasks, not just those which require more (vs. less) information processing. The potential effects of other forms of task complexity on locomotors' and assessors' performance will be returned to in the General Discussion section. 2. Study 1 The goal of Study 1 was twofold: (1) to investigate our hypotheses about the interactive effect of locomotion and assessment on task performance in tasks of varied complexity, and (2) to examine these hypotheses in a real-world organizational setting. 2.1. Participants Ninety-four employees (49 women) from an Italian organization (a supermarket), divided into 9 work units, participated in this research on a voluntary basis. They did not receive any compensation in exchange for their participation. Employees' mean age was 35.35 years (SD = 10.52; Min = 19; Max = 66), their mean job tenure was 7.48 years (SD = 7.44; Min = 1; Max = 40), and the mean group size was 13.04 (SD = 4.47; Min = 3; Max = 18). Most of the participants (60.6%) had a high school diploma, 26.5% had a junior high school diploma, and 6.6% had a college degree. Participants were informed about the study and consented in writing to the use of their anonymized data. The study complied with the Declaration of Helsinki and was approved by the local ethics committee. 2.2. Procedure Participants filled out the Regulatory Mode Scale (Kruglanski et al., 2000) to measure their locomotion and assessment tendencies. This was followed by the Job Complexity Scale, a 4 item self-report scale designed to measure the amount of information processing needed at work. The major dependent variable was a measure of subordinates' performance as rated by their supervisors.
2.3. Materials 2.3.1. Locomotion and assessment Kruglanski et al. (2000) developed two scales designed to measure the individual difference tendencies of locomotion and assessment. Analyses using several samples from the United States and Italy revealed that the two scales were temporally stable and internally consistent. In addition, validation studies have shown that the two constructs are relatively independent and have strong predictive validity, as well as good convergent and divergent validity (Kruglanski et al., 2000). The Italian version of the locomotion and assessment scales (Kruglanski et al., 2000) constitutes two separate 12-item self-report measures designed to tap individual differences in these tendencies. Specifically, respondents rate the extent to which they agree with self-descriptive statements reflecting locomotion (e.g., “By the time I accomplish a task, I already have the next one in mind”) and assessment (e.g., “I spend a great deal of time taking inventory of my positive and negative characteristics”). Ratings are made on a 6-point Likert-type scale with the response alternatives anchored at the ends with 1 (strongly disagree) to 6 (strongly agree). We computed two composite scores (one for locomotion and one for assessment) by averaging across responses to each locomotion and assessment item, respectively. In this sample, the Cronbach's α was .74 for both the locomotion and assessment scales.
2.3.2. Job complexity Following prior research on the topic (e.g., Campbell, 1988), in this study we operationalized job complexity as the amount of information processing needed at work. To assess the degree to which their jobs required workers to attend to and process data or other information, we used the 4-item Information Processing scale. The Information Processing scale is a subscale of the larger Work Design Questionnaire, which was previously validated by Morgeson and Humphrey (2006). The items presented to participants were: (1) “The job requires me to monitor a great deal of information,” (2) “The job requires that I engage in a large amount of thinking,” (3) “The job requires me to keep track of more than one thing at a time,” and (4) “The job requires me to analyze a lot of information.” Items were answered on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). Cronbach's α for the four items in the present study was .63.
2.3.3. Performance Participants' performance was measured by supervisors' ratings of them on a 4 item instrument (adapted from Gould, 1979; Hobman, Jackson, Jimmieson, & Martin, 2011; Pierro, Pica, Kruglanski, & Higgins, 2014). Items included an assessment of quantity of performance, quality of performance, and an overall rating of performance. Specifically, we asked supervisors to appraise the performance of their employees (i.e., the research participants) over the last year1 by responding to the following items: (1) “In terms of percentage, to what extent were the employee's objectives reached during the last year?”, (2) “How do you evaluate the employee's overall performance in the last year?”, (3) “How do you evaluate the quality of the work he/she has done in the last year?”, and (4) “How do you evaluate the quantity or volume of the work he/she has done in the last year?” The first item was rated on a 10-point scale ranging from 1 (10%) to 10 (100%), and the last three items were rated on 10-point scales ranging from 1 (extremely negative) to 10 (extremely positive). We computed a composite performance score by averaging across the responses to each item (Cronbach's α = .97).
1
Note that, as mentioned above, the minimum of participants' job tenure was 1 year.
M. Chernikova et al. / Personality and Individual Differences 89 (2016) 134–142 Table 1 Descriptive statistics and variable intercorrelations for Study 1.
1. Locomotion 2. Assessment 3. Job complexity 4. Performance
M (SD)
1
2
3
4
4.89 (0.51) 3.10 (0.74) 3.89 (0.60) 7.78 (1.28)
– .03 −.13 .15
– .13 −.34⁎⁎
– −.10
–
N = 94. ⁎⁎ p b .01.
3. Results Table 1 contains a summary of descriptive statistics and zero-order correlations between the variables. The predictions regarding the effect on performance of interaction between regulatory mode orientations and job complexity were tested by means of a moderated multiple regression analysis. We used the PROCESS program for this analysis (Model 3, Hayes, 2013). In this moderated multiple regression analysis the main effects of locomotion, assessment, and job complexity, and all possible two-way interactions and the three-way interaction were entered. All variables were standardized and the interaction terms were based on these standardized scores. The main effects of gender, age, job tenure, and education were also entered as control variables. The overall regression was significant (R2 = .40, F(11, 82) = 5.05, p b .001); the results of this analysis are reported in Table 2. As can be seen in Table 2, the results show a main effect of locomotion on performance, with higher locomotion associated with greater performance. In line with our predictions, there is also a positive and significant effect of the interaction between locomotion and assessment on performance. Of greatest importance for the current hypothesis, the three-way interaction is significant. The positive sign of the three-way interaction suggests that, as predicted, the complementarity effect is stronger in high complexity jobs (Hypothesis 2). These findings are illustrated via the predicted mean values shown in Fig. 1a and b. Following the suggestion of Aiken and West (1991), the values were obtained using one standard deviation above and below the means of the relevant variables in the regression equation. The simple interaction analysis conducted to further understand the nature of the three-way interaction (Aiken & West, 1991) revealed that the interaction effect between locomotion and assessment, in predicting performance, was significant in the high complexity job (1 SD above the mean: b = .89, SE = .24, t = 3.68, p b .001), but not significant in the low complexity job (1 SD below the mean: b = −.25, SE = .24, t = −1.05, p = .298). Moreover, the simple slope analysis conducted to examine the twoway interaction (locomotion × assessment) only for those participants in the high complexity job (1 SD above the mean) revealed that the relationship between locomotion and performance was positive and significant for participants relatively high in assessment (1 SD above the mean: b = 1.30, SE = .37, t = 3.47, p b .001), whereas this relationship was negative and significant for participants relatively low in assessment (1 SD below the mean: b = −.47, SE = .19, t = − 2.19, p =
Table 2 Summary of moderated regression analysis in Study 1.
Locomotion Assessment Complexity Loc × Ass Loc × Complex Ass × Complex Loc × Ass × Complex Age Gender Education Tenure
b
Beta
SE
t
p
.31 −.22 .18 .32 .10 .20 .57 .00 −.30 −.09 .05
.24 −.17 .14 .20 .07 .19 .32 .01 .12 −.04 .33
.13 .12 .13 .15 .16 .11 .19 .01 .24 .21 .02
2.42 −1.82 1.38 2.09 .63 1.86 3.02 .09 −1.25 −.45 2.65
.017 .072 .171 .040 .529 .066 .003 .926 .214 .656 .001
137
.032). On the other hand, the relationship between assessment and performance was positive and significant for participants relatively high on locomotion (1 SD above the mean: b = .86, SE = .33, t = 2.56, p = .012), whereas this relationship was negative and significant for participants relatively low in locomotion (1 SD below the mean: b = −.91, SE = .26, t = − 3.52, p b .001). These results provide support for Hypothesis 2. Although the simple interaction was not significant in low complexity jobs, a simple slope analysis revealed that the simple slope is in the direction that we would predict, such that the relationship between assessment and performance was negative and significant for participants relatively high in locomotion (1 SD above the mean; b = −.68, SE = .25, t = −2.74, p = .007), whereas this relationship was not significant for participants relatively low in locomotion (1 SD below the mean; p = .59). On the other hand, the relationship between locomotion and performance was not significant for participants relatively high in assessment (1 SD above the mean; p = .91), whereas this relationship was positive and trending toward significance for participants relatively low in assessment (1 SD below the mean; b = .47, SE = .29, t = 1.64, p = .105). These results provide support for Hypothesis 1. We acknowledge that the structure of the data is nested (i.e., individual performance ratings are nested within units) and that this may raise the concern of non-independent data. Thus, to estimate how much of the total variance in performance ratings was betweenunit, we calculated the intraclass correlation coefficient (ICC). The result yielded a coefficient of .13, suggesting that only a small proportion of the variance in performance ratings was between unit. This ICC value close to zero implies the independence of the data and suggests that the ordinary least squares (OLS) approach used above can be sufficient. However, consistent with research (for a brief review, see Peugh, 2010) that has shown ICC values between .05 and .20 to be common in multilevel modeling applications in social science research, and following the recommendation of Hayes (2006), we also applied a multilevel modeling approach to the data, using maximum likelihood (ML) estimation. In the analysis, reported in Table 3, we entered as fixed all our level-one control variables (gender, age, education, and job tenure) and the main predictor variables (locomotion, assessment, job complexity, and the interaction between them). Locomotion, assessment, and job complexity were grand mean centered, and the interaction terms was based on these centered scores. We added the group size (level-two or group level variable) as a fixed control variable. Finally, only the intercept was a random effect, entered at the unit (group) level. Focusing our attention on the fixed effects, the results of this analysis revealed that group size was not related to performance. More important, consistent with above reported results, the analysis confirms: (1) a significant and positive main effect of locomotion on performance; (2) a significant and positive two-way interaction effect between locomotion and assessment; and, of greatest importance, (3) the significant and positive 3-way interaction effect between locomotion, assessment, and job complexity. Finally, the random effect was not significant (B = .03, SE = .05, Wald = .62, p = .533), providing no support for the notion that the intercepts varied between units. These findings effectively show that our conclusions here are not compromised by the potential dependency of observations. 3.1. Discussion The results of this study provide initial support for both of our hypotheses: when job tasks involved less complexity, those who were high on locomotion and low on assessment had the best performance (Hypothesis 1); on the other hand, when job tasks involved more complexity, individuals who were high on both locomotion and assessment had the highest performance (Hypothesis 2). It should be noted that since the results within low job complexity were only trending toward significance, this study alone does not allow us to draw firm conclusions regarding the influence of regulatory mode on performance under
138
M. Chernikova et al. / Personality and Individual Differences 89 (2016) 134–142
Fig. 1. a. Interactive effect of locomotion and assessment on job performance under low job complexity (Study 1). b. Interactive effect of locomotion and assessment on job performance under high job complexity (Study 1).
conditions of low job complexity. In addition, the first study was correlational in nature and therefore mute with respect to causality. Since work performance was measured retroactively (i.e., supervisors' ratings of participants' job performance over the last year), it is possible that participants' work performance may have influenced their regulatory mode levels. The next study was conducted to address these issues. 4. Study 2 The major goal of Study 2 was to extend the findings of the first study with the use of an experimental manipulation of task complexity. In addition, we used a different measure of performance: participants' number of correct responses in an inbox organization task. Table 3 Multilevel analyses.
Locomotion Assessment Complexity Loc × Ass Loc × Complex Ass × Complex Loc × Ass × Complex Age Gender Education Tenure Group size
b
SE
t
p
.27 −.22 .21 .30 .06 .21 .49 .00 −.22 −.10 .05 −.02
.12 .11 .12 .14 .15 .10 .18 .01 .23 .20 .02 .03
2.23 −1.93 1.70 2.11 .40 2.14 2.72 .18 −.92 −.49 2.78 −.67
.028 .056 .091 .037 .689 .035 .008 .858 .359 .626 .007 .507
4.1. Participants One hundred and three undergraduate participants (57 females) completed the study at Sapienza University of Rome in exchange for partial course credit. The average age of participants was 23.70 (SD = 2.83). There were no effects of gender or age, so they will not be discussed further. Participants were informed about the study and consented in writing to the use of their anonymized data. The study complied with the Declaration of Helsinki and was approved by the local ethics committee.
4.2. Procedure Participants filled out the Regulatory Mode Scale (Kruglanski et al., 2000) to measure their locomotion and assessment tendencies. They subsequently came into the lab and completed either a simple or complex inbox task. The major dependent variable was their performance on the inbox task.
4.3. Materials 4.3.1. Locomotion and assessment Participants completed the same regulatory mode questionnaire used in Study 1 in a mass testing session one month prior to coming into the lab. In this study, the Cronbach's α was .80 for the locomotion scale and .78 for the assessment scale. The two scales were not
M. Chernikova et al. / Personality and Individual Differences 89 (2016) 134–142
correlated in this sample, consistent with findings from previous studies (Kruglanski et al., 2000). 4.3.2. Task complexity Upon arriving in the lab, participants completed an inbox task that involved organizing and responding to various emails (previously used in Jimmieson & Terry, 1997; Jimmieson & Terry, 1999; Parker, Jimmieson, & Amiot, 2009; Parker, Jimmieson, & Amiot, 2013). All participants received a booklet containing instructions for the task and the pieces of information they needed to read and respond to (e.g., a team performance review and an organizational chart). Participants were instructed to assume the role of Alex Jennings, an Interim Retail Manager in a branch of the Borough Bank. They had to manage two team leaders and attempt to improve the branch's performance. In order to do this, they had to read a number of items (e.g., documents, emails, and memos) about human resources and financial issues. Participants had to write out the action they would take with regard to each of the items. They were told they should list as many actions as they considered necessary. Participants were randomly assigned to either the simple condition, in which they had to read and respond to six items (N = 52); or the complex condition, in which they had to read and respond to twelve items (N = 51). Independently of the task condition they were assigned to, all participants had a total of 90 min to complete the task. 4.3.3. Manipulation check After the inbox task, all participants filled out a 5 item manipulation check questionnaire of perceived task complexity (Cronbach's α = .80). The questionnaire contained one item measuring overall task complexity (“The task I have done is complex”) and 4 items measuring the amount of information required to process the task (e.g., “The task required me to manage a large amount of information”). After completing the manipulation check questionnaire, participants were provided with feedback on their performance on the inbox task; they were then thanked for their participation and dismissed. 4.4. Results We performed an ANOVA on the perceived complexity measure, which revealed that our manipulation of complexity was successful (F(1, 101) = 10.61, p = .002): participants in the simple condition reported lower levels of perceived complexity (M = 3.66, SD = 1.08) than those who were in the complex condition (M = 4.28, SD = .85). In addition, an ANOVA on the time measure (F(1, 101) = 28.09, p b .001) revealed that participants in the simple condition needed less time to complete the task (M = 39.04 min, SD = 20.20) than those in the complex condition (M = 62.37 min, SD = 24.33). This supports the notion that the simple task involved less information processing, as greater information processing is theorized to be more timeconsuming (Robinson, 2001). Performance on the inbox task was calculated by computing each participant's mean amount of correct actions per item. Whether the answers were marked as correct or not was established following a codebook written by the creators of the task (https://www.assessmentday. co.uk/answers-in-tray-assessmentday.pdf), which has been used in other published work on the inbox task (e.g., Jimmieson & Terry, 1997). For instance, one of the items participants had to address was an email from their subordinate, a sales executive, informing them that he was resigning due to a lack of developmental opportunities at the firm. Correct responses to this item included: (1) setting a meeting with the sales executive's direct supervisor to investigate any problems he may have had, (2) emailing the HR manager to gain an understanding of any issues that may have led to the resignation, and (3) assuring the staff that they will be offered more developmental opportunities from now on. Responses were scored for correctness by a trained coder who used the codebook, and who was blind to participants' locomotion and assessment scores.
139
Table 4 Descriptive statistics and variable intercorrelations for Study 2.
1. Locomotion 2. Assessment 3. Performance
M (SD)
1
2
3
4.42 (0.64) 3.84 (0.70) 1.60 (0.66)
– .07 .14
– −.02
–
N = 103.
Our predictions regarding the effect on task performance of the interaction between locomotion, assessment, and complexity condition were tested with a moderated multiple regression analysis. We used the PROCESS program for this analysis (Model 3; Hayes, 2013). Table 4 contains a summary of descriptive statistics and zero-order correlations between the variables. In the moderated multiple regression analysis the main effects of locomotion, assessment, and complexity condition, all possible two-way interactions, and the three-way interaction were entered. Following Aiken and West's (1991) recommendation, locomotion and assessment scores were mean centered and the interaction terms were based on these centered scores; task complexity condition was contrast coded as −1 (simple condition) and 1 (complex condition). The main effect of time was entered as a control variable, since participants differed in the amount of time they spent on the task in the simple and complex task conditions. Results of the moderated multiple regression analysis are reported in Table 5. The overall model was significant (R2 = .31, F(8, 94) = 5.40, p b .001). There was a main effect of condition, such that participants in the simple condition had better performance than those in the complex condition. There was also a main effect of time, such that the more time participants spent on the task, the better their performance. No other main effects were significant, nor were there any significant two-way interactions. Most importantly, as predicted, the main effects we observed were qualified by a significant three-way interaction (see Fig. 2a & b for interaction graphs). The simple interaction analysis conducted to further understand the nature of the three-way interaction (Aiken & West, 1991) revealed that the interaction effect between locomotion and assessment in predicting task performance was trending toward significance in both the simple condition (B = −.14, SE = .07, t(94) = − 1.96, p = .052) and the complex condition (B = .17, SE = .10, t(94) = 1.67, p = .097). The simple slope analyses revealed that when assessment was low in the simple task condition, high locomotors performed significantly better than low locomotors (B = .27, SE = .13, t(94) = 2.10, p = .038). When assessment was high in the simple condition, there were no significant differences between high and low locomotors (p = .76). These results support Hypothesis 1. When assessment was low in the complex task condition, there were no significant differences between high and low locomotors (p = .47). When assessment was high in the complex condition, high locomotors performed significantly better than low locomotors (B = .24, SE = .12, t(94) = 2.00, p = .048). These results support Hypothesis 2. When locomotion was low in the simple task condition, there were no significant differences between high and low assessors (p = .97). Table 5 Summary of moderated regression analysis in Study 2.
Locomotion Assessment Complexity cond. Loc × Ass Loc × Complex cond. Ass × Complex cond. Loc × Ass × Complex cond. Time
B
Beta
SE
t
p
.10 −.02 −.23 .00 −.02 .12 .16 .01
.14 −.03 −.34 .01 −.03 .18 .23 .51
.06 .06 .07 .06 .06 .06 .06 .00
1.65 −.37 −3.34 .15 −.39 1.97 2.51 5.21
.103 .714 .001 .884 .697 .052 .014 .000
140
M. Chernikova et al. / Personality and Individual Differences 89 (2016) 134–142
Fig. 2. a. Interactive effect of locomotion and assessment on performance in the simple task condition (Study 2). b. Interactive effect of locomotion and assessment on performance in the complex task condition (Study 2).
When locomotion was high in the simple condition, high assessors performed significantly worse than low assessors (B = −.30, SE = .12, t(94) = −2.56, p = .012). These results support Hypothesis 1. When locomotion was low in the complex task condition, there were no significant differences between high and low assessors (p = .616). When locomotion was high in the complex condition, there was a trend such that high assessors performed better than low assessors (B = .27, SE = .15, t(94) = 1.81, p = .072). These results support Hypothesis 2. 4.5. Discussion In summary, in the second study we found that high locomotion and low assessment led to better performance when a task was simple (supporting Hypothesis 1), whereas high locomotion and high assessment led to better performance when a task was complex (supporting Hypothesis 2). Importantly, although this study used a different task and a different measure of task performance, the results mirrored those from the first study. This suggests that our findings are not method-specific and can be extended to a variety of contexts. 5. General discussion Across two studies, we found that individuals high on locomotion and low on assessment performed better on simple tasks, whereas individuals high on both locomotion and assessment performed better on complex tasks. In the first study, when the work of employees in an Italian organization involved more information processing, those who were high on both locomotion and assessment had the best performance; on the other hand, when their work involved less information processing, those who were high on locomotion but low in assessment
had the highest performance. In the second study, university students who completed a simple inbox task in the lab performed best when they were high on locomotion and low on assessment. In contrast, students who performed a complex inbox task performed best when they were high on both locomotion and assessment. Taken together, the two studies suggest that high levels of both locomotion and assessment are necessary to successfully perform complex—but not simple—tasks. For simple tasks, high levels of locomotion alone are enough to ensure successful performance. To return to the story of Michael and Max, it seems that contrary to intuition, Max will not always be the best performer. Though the combination of careful thought and speedy action can often lead to high performance, there are some tasks which benefit most from a focus on swift action—and as such, individuals high on the trait of locomotion will have a significant advantage on such tasks. Thus, Max and Michael are each well-suited to performing tasks of different complexity levels, and would do well to seek out cognitive tasks that fit well with their regulatory mode orientation. The present studies extend the literature on regulatory mode in important ways. Most prior research on the interactive effects of locomotion and assessment has suggested that the high-high combination of locomotion and assessment is generally beneficial for both individual and team performance (e.g., Mauro et al., 2009; Pierro, Pica, Mauro, Kruglanski and Higgins, 2012; Pierro, Presaghi, Higgins, Klein and Kruglanski, 2012; see Kruglanski et al., 2013, for a review). However, contrary to prior research and theorizing on regulatory mode (e.g., Kruglanski et al., 2013), our studies show that being high on both locomotion and assessment may not always be advantageous for goal pursuit. Rather, differing levels of task complexity can determine whether an individual who is high only on locomotion, only on
M. Chernikova et al. / Personality and Individual Differences 89 (2016) 134–142
assessment, or high on both regulatory modes will have the highest performance on a task. Thus the current studies offer an important insight into the limits of the beneficial effects of a high-high regulatory mode combination. They also suggest that future theorizing on regulatory mode should explore the important question of which circumstances are most conducive to the performance of individuals who are high on assessment, high on locomotion, or high on both. This research can have intriguing implications for employee performance in organizations. An employer can evaluate the complexity of the tasks that a particular job requires and subsequently select employees based on their regulatory mode orientations. Individuals who are high on locomotion but low in assessment should be selected for tasks that are relatively simple, while those who are high on both locomotion and assessment should be selected for relatively complex tasks. The findings from these studies can also be extended to dyadic and group contexts: for instance, if two individuals are working together and are both high on locomotion, they should perform best on simple tasks. On the other hand, if one of the individuals is high on assessment and the other is high on locomotion, they should perform best on complex tasks. One limitation of these studies is that locomotion and assessment were measured and not manipulated in both studies; it would be important for this to be addressed in future research on the effects of locomotion and assessment complementarity. Another limitation is that both of our studies operationalize task complexity as workload, or the amount of information an individual has to process. While this operationalization of cognitive task complexity is supported by previous literature on complexity, there are also many other operationalizations of complexity which were not addressed in our studies (e.g., uncertainty of outcomes, multiple alternative paths to goal completion, and rate of information change, among others; Campbell, 1988). Given the lack of multiple operationalizations of complexity in the current studies, future research would do well to investigate the current hypotheses using different operationalizations of simple and complex tasks. A third limitation of this research is that both of our studies focused on cognitive task complexity as opposed to other types of complexity; as a result, we cannot speak to whether our results are generalizable to noncognitive tasks. Future studies could examine whether performance on tasks of varying physical complexity (e.g., assuming an easy vs. difficult yoga position) can also be influenced by individuals' regulatory mode. Other future directions for this line of research include investigating the interactive effect of regulatory mode complementarity and task complexity on outcomes besides performance. For instance, in addition to their higher performance on simple tasks, individuals who are high on locomotion and low on assessment may have higher levels of wellbeing and satisfaction when they work on simple tasks. On the other hand, individuals who are high on both locomotion and assessment may have higher levels of well-being and satisfaction when they work on complex tasks. In the organizational realm, when individuals experience a fit between their regulatory mode and the level of complexity their job requires, they may invest more effort in their job, exhibit fewer withdrawal behaviors (e.g., absenteeism, lateness), and cope better with organizational change. Future research can also examine other variables that moderate the effects of locomotion and assessment on performance: for example, high locomotors may perform better under time pressure than either high assessors or those who are high on both locomotion and assessment. Lastly, an anonymous reviewer made the insightful point that cognitive tasks can differ in information processing requirements in terms of whether those requirements pertain to planning (where planning requires deep thinking, but then execution does not) or execution (where planning is not feasible or helpful, and thus not complex, but execution requires complex information processing). Though the design of our current studies does not allow us to disentangle which of these aspects of task complexity was driving our effects, future research could separately manipulate increased
141
complexity during the planning stage, execution stage, or both stages of a task. It may be that high assessors are better at tasks that involve more complexity during the planning stage, high locomotors are better at tasks that involve more complexity at the execution stage, and individuals who are high on both locomotion and assessment are better at tasks which involve complexity during both the planning and execution stage. In conclusion, by elucidating the conditions under which various combinations of locomotion and assessment lead to higher task performance, the current studies offer a compelling addition to existing research on the effects of regulatory mode. These findings can have intriguing applications for organizations and other real-world contexts. They also provide support for the notion that regulatory mode is a useful and important social psychological construct, one that is relevant across many domains of research. References Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage. Amato, C., Pierro, A., Chirumbolo, A., & Pica, G. (2014). Regulatory modes and time management: How locomotors and assessors plan and perceive time. International Journal of Psychology, 49(3), 192–199. http://dx.doi.org/10.1002/ijop.12047. Avnet, T., & Higgins, E. T. (2003). Locomotion, assessment, and regulatory fit: Value transfer from “how” to “what”. Journal of Experimental Social Psychology, 39(5), 525–530. Campbell, D. J. (1988). Task complexity: A review and analysis. Academy of Management Review, 131(1), 40–52. Gould, S. (1979). Age, job complexity, satisfaction, and performance. Journal of Vocational Behavior, 14, 209–223. Hayes, A. F. (2006). A primer on multilevel modeling. Human Communication Research, 32(4), 385–410. Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York, NY: Guilford Press. Higgins, E. T., Kruglanski, A. W., & Pierro, A. (2003). Regulatory mode: Locomotion and assessment as distinct orientations. In M. P. Zanna (Ed.), Advances in experimental social psychology. 35. (pp. 293–344). New York: Academic Press. Hobman, E. V., Jackson, C. J., Jimmieson, N. L., & Martin, R. (2011). The effects of transformational leadership behaviours on follower outcomes: An identity-based analysis. European Journal of Work and Organizational Psychology, 20(4), 553–580. Jimmieson, N. L., & Terry, D. J. (1997). Responses to an in-basket activity: The role of work stress, behavioral control, and informational control. Journal of Occupational Health Psychology, 2, 72–83. Jimmieson, N. L., & Terry, D. J. (1999). The moderating role of task characteristics in determining responses to a stressful work simulation. Journal of Organizational Behavior, 20, 709–736. Kruglanski, A. W., Pierro, A., Mannetti, L., & Higgins, T. E. (2013). The distinct psychologies of “looking” and “leaping”: Assessment and locomotion as the springs of action. Social and Personality Psychology Compass, 7(2), 79–92. Kruglanski, A. W., Thompson, E. P., Higgins, E. T., Atash, M. N., Pierro, A., Shah, J. Y., & Spiegel, S. (2000). To do the right thing or to just do it!: Locomotion and assessment as distinct self-regulatory imperatives. Journal of Personality and Social Psychology, 79(5), 793–815. Kruglanski, A. W., Pierro, A., & Higgins, E. T. (2015). Experience of time by people on the go: A theory of the locomotion-temporality interface. Personality and Social Psychology Review, 1-18. Mannetti, L., Pierro, A., Higgins, E. T., & Kruglanski, A. W. (2012). Maintaining physical exercise: How locomotion mode moderates the full attitude–intention–behavior relation. Basic and Applied Social Psychology, 34(4), 295–303. March, J., & Simon, H. (1958). Organizations. New York: Wiley. Mauro, R., Pierro, A., Mannetti, L., Higgins, E. T., & Kruglanski, A. W. (2009). The perfect mix: Regulatory complementarity and the speed–accuracy balance in group performance. Psychological Science, 20(6), 681–685. Morgeson, F. P., & Humphrey, S. E. (2006). The Work Design Questionnaire (WDQ): Developing and validating a comprehensive measure for assessing job design and the nature of work. Journal of Applied Psychology, 91(6), 1321. Orehek, E., & Vazeou-Nieuwenhuis, A. (2013). Sequential and concurrent strategies of multiple goal pursuit. Review of General Psychology, 17(3), 339. http://dx.doi.org/10. 1037/a0032584. Orehek, E., Mauro, R., Kruglanski, A. W., & van der Bles, A. M. (2012). Prioritizing association strength versus value: The influence of self-regulatory modes on means evaluation in single goal and multigoal contexts. Journal of Personality and Social Psychology, 102(1), 22–31. http://dx.doi.org/10.1037/a0025881. Parker, S. L., Jimmieson, N. L., & Amiot, C. E. (2009). The stress-buffering effects of control on task satisfaction and perceived goal attainment: An experimental study of the moderating influence of desire for control. Applied Psychology: An International Review, 58, 622–652. Parker, S. L., Jimmieson, N. L., & Amiot, C. E. (2013). Self-determination, control, and reactions to changes in workload: A work simulation. Journal of Occupational Health Psychology, 18(2), 173. Peugh, J. L. (2010). A practical guide to multilevel modeling. Journal of School Psychology, 48(1), 85–112.
142
M. Chernikova et al. / Personality and Individual Differences 89 (2016) 134–142
Pierro, A., Pica, G., Mauro, R., Kruglanski, A. W., & Higgins, E. T. (2012a). How regulatory modes work together: Locomotion-assessment complementarity in work performance. Testing, Psychometrics, Methodology in Applied Psychology, 19, 247–262. Pierro, A., Presaghi, F., Higgins, E. T., Klein, K. M., & Kruglanski, A. W. (2012b). Frogs and ponds: A multilevel analysis of the regulatory mode complementarity hypothesis. Personality and Social Psychology Bulletin, 38(2), 269–279. Pierro, A., Giacomantonio, M., Pica, G., Kruglanski, A. W., & Higgins, E. T. (2011). On the psychology of time in action: Regulatory mode orientations and procrastination. Journal of Personality and Social Psychology, 101(6), 1317. Pierro, A., Pica, G., Kruglanski, A. W., & Higgins, T. E. (2014). Regulatory mode orientations and self-forgiveness. Unpublished manuscript, University of Rome “La Sapienza,” Rome, Italy.
Robinson, P. (2001). Task complexity, task difficulty, and task production: Exploring interactions in a componential framework. Applied Linguistics, 22(1), 27–57. Schroder, H., Driver, M., & Streufert, S. (1967). Human information processing. New York: Holt, Rinehart and Winston. Schwab, D., & Cummings, L. (1976). A theoretical analysis of the impact of task scope on employee performance. Academy of Management Review, 1, 23–35. Terborg, J., & Miller, H. (1978). Motivation, behavior and performance: A closer examination of goal-setting and monetary incentives. Journal of Applied Psychology, 63, 29–39.
Contents lists available at ScienceDirect
Personality and Individual Differences journal homepage: www.elsevier.com/locate/paid
Different strokes for different folks: Effects of regulatory mode complementarity and task complexity on performance☆ Marina Chernikova a,⁎, Calogero Lo Destro b, Romina Mauro b, Antonio Pierro b, Arie W. Kruglanski a, E. Tory Higgins c a b c
University of Maryland, United States Universita di Roma “La Sapienza”, Italy Columbia University, United States
a r t i c l e
i n f o
Article history: Received 21 July 2015 Received in revised form 1 October 2015 Accepted 3 October 2015 Available online 22 October 2015 Keywords: Self-regulation Locomotion Assessment Regulatory mode Performance
a b s t r a c t We examine how task features interact with individuals' regulatory modes in determining performance. Effective goal pursuit normally occurs when high locomotion (the regulatory mode concerned with motion from state to state) and high assessment (the regulatory mode concerned with critical evaluation) work together (Kruglanski et al., 2000). However, there may be situations in which this high–high combination is unnecessary or even detrimental to good performance. We hypothesized that on simple tasks, high locomotion and low assessment should lead to the best performance; on complex tasks, however, high locomotion and high assessment should lead to the best performance. We tested these hypotheses in two empirical studies, one carried out in an organizational setting, the other in the lab. In the first study, we measured individuals' locomotion and assessment tendencies, asked them to rate the complexity of their daily work tasks, and obtained measures of their job performance from their supervisors. In the second study, we measured individuals' locomotion and assessment tendencies, manipulated the task complexity of an inbox task they had to complete, and measured their performance on that task. Both studies provided support for our hypotheses. These results offer important insights regarding the effects of regulatory mode on performance. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Consider a brief tale of two friends. One of them, Michael, is always energetic and quick at his work. The other friend, Max, is also energetic and quick; in addition to this, Max also likes to methodically analyze his options and decide upon the best course of action before he starts. When they are given a task to complete, Michael begins right away and speeds through the task; Max, on the other hand, spends some time thinking deeply about the task and how to best complete it, then also gets going and completes the task quickly. It seems fairly straightforward to assume that Max (who focuses on both careful thought and swift action) will generally outperform Michael (who focuses only on swift action). But is this intuition correct? In order to answer this question, it is essential to distinguish between two basic aspects of self-regulation: locomotion and assessment. Regulatory mode theory postulates that every act of self-regulation involves two components: evaluating which goal or means to select during goal pursuit, and committing energy to engaging in the chosen ☆ We would like to thank Katarzyna Jasko and Maxim Milyavsky for their valuable comments on an earlier draft of the paper. ⁎ Corresponding author. E-mail address: [email protected] (M. Chernikova).
http://dx.doi.org/10.1016/j.paid.2015.10.011 0191-8869/© 2015 Elsevier Ltd. All rights reserved.
option (Higgins, Kruglanski, & Pierro, 2003; Kruglanski et al., 2000). These motivational components, or regulatory modes, are referred to as assessment and locomotion: assessment is concerned with comparison and evaluation, and locomotion is concerned with making progress toward a goal. The two regulatory modes can be measured as individual traits (Kruglanski et al., 2000) or manipulated as state variables (Avnet & Higgins, 2003). Furthermore, locomotion and assessment are assumed to be functionally independent, so that an individual can be high on both, low on both, or high on one and low on the other (Higgins et al., 2003; Kruglanski et al., 2000). An individual's regulatory mode can influence how she carries out a wide variety of activities in her daily life, ranging from the types of goals she selects (e.g.Mannetti, Pierro, Higgins, & Kruglanski, 2012; Orehek, Mauro, Kruglanski, & van der Bles, 2012) to the manner in which she pursues those goals (e.g. Avnet & Higgins, 2003; Orehek & Vazeou-Nieuwenhuis, 2013). The implications of each regulatory mode for goal pursuit will be elaborated in the following sections. Locomotion is defined as “the aspect of self-regulation concerned with movement from state to state, and with committing the psychological resources that will initiate and maintain goal-related movement in a straightforward and direct manner, without undue distractions or delays” (Kruglanski et al., 2000, p. 794). As such, the essential motivation of individuals high on locomotion is to experience psychological
M. Chernikova et al. / Personality and Individual Differences 89 (2016) 134–142
movement. High locomotors emphasize “hurrying,” “doing,” and “getting on with it,” since all of those allow them to avoid standing still and doing nothing (Higgins et al., 2003). Previous research suggests that the locomotion orientation can influence many aspects of goal pursuit. For example, high locomotors avoid procrastinating on goals and tasks: locomotion is negatively associated both with scores on a procrastination scale (which measures the tendency to delay task initiation or completion), and with insurance workers' actual procrastination as measured over a three month period (Pierro, Giacomantonio, Pica, Kruglanski, & Higgins, 2011). High locomotors tend to complete tasks as quickly as possible (even when such a focus on speed comes at the expense of accuracy; Kruglanski et al., 2000; Mauro, Pierro, Mannetti, Higgins, & Kruglanski, 2009). For instance, when engaged in a proofreading task, high (vs. low) locomotors take significantly less time to finish the task (Kruglanski et al., 2000). High locomotors exhibit a greater ability to stay focused on a task and avoid becoming distracted: locomotion is positively associated both with perseverance (as measured by a scale tapping the ability to sustain effort in the face of diversity) and resistance to temptation (as measured by participants' reports of how likely they would be to put off studying for an important exam; Pierro et al., 2011). Individuals who are high on locomotion are better at managing the time they devote to their goals: high locomotors have an increased proficiency at setting goals and priorities, a greater preference for organization, and greater perceived control over their time (Amato, Pierro, Chirumbolo, & Pica, 2014). High locomotors are more likely to select high expectancy (vs. high value) options during goal pursuit; in other words, locomotors tend to select means which serve only one goal (and thus may have a greater expectancy of attaining that goal) rather than means which serve multiple goals (and may have a lower expectancy of attaining any one of them; Orehek et al., 2012). Individuals who are high (vs. low) on locomotion are more likely to follow through on initial goals they set (such as attending fitness classes; Mannetti et al., 2012). Lastly, when selecting a means to their goal, high locomotors prefer to make simultaneous evaluations of all possible alternatives (rather than engaging in a sequential series of comparisons; Avnet & Higgins, 2003). Thus, locomotion can influence goal pursuit in a wide variety of ways. Assessment is defined as “the comparative aspect of self-regulation concerned with critically evaluating entities or states, such as goals or means, in relation to alternatives in order to judge relative quality” (Kruglanski et al., 2000, p. 794). The essential motivation of individuals high on assessment is to “make the right choice”; as such, high assessors are preoccupied with making comparisons and critical evaluations to ensure that they arrive at the correct decision before moving forward (Higgins et al., 2003). Like locomotion, assessment can also impact various aspects of individuals' goal pursuit. For instance, individuals who are high on assessment tend to procrastinate on goals and tasks: assessment is positively correlated both with scores on a procrastination scale, and with insurance workers' actual procrastination as measured over a three month period (Pierro et al., 2011). High assessors are slower to complete tasks, but are also more accurate at those tasks (Kruglanski et al., 2000; Mauro et al., 2009). For example, high (vs. low) assessors were found to be significantly more accurate when completing a proofreading task (Kruglanski et al., 2000). High assessors also tend to have greater concern over potential mistakes during goal pursuit (measured with items such as “I should be upset if I make a mistake”), higher personal standards for their task performance (measured with items such as “I expect higher performance in my daily tasks than most people”), and greater doubts about whether they are making the right choice (measured with items such as “I tend to get behind in my work because I repeat things over and over”; Pierro et al., 2011). High assessors are more likely to select high value (vs. high expectancy) options during goal pursuit: they tend to choose means which serve multiple goals (and can thus produce more overall value in terms of the amount of goals they attain) rather than means which serve only one goal (and
135
which therefore produce less overall value; Orehek et al., 2012). High assessors have a greater tendency to form behavioral goals (such as the goal to attend gym classes) based on perceived social norms (Mannetti et al., 2012). Lastly, high assessors prefer to select a means to their goal by making a sequential series of comparisons among the possible alternatives (rather than simultaneously evaluating all their options; Avnet & Higgins, 2003). In summary, previous research has shown that both assessment and locomotion regulatory mode are relevant to goal pursuit in a variety of domains. As described earlier, regulatory mode theory posits that locomotion and assessment are independent; therefore, if an individual is low on one of the orientations, he will not necessarily be high on the other (Higgins et al., 2003; Kruglanski et al., 2000). This independence assumption implies that different combinations of locomotion and assessment can have very different effects on goal pursuit and attainment. Prior research on the interactive effects of locomotion and assessment has generally focused on the beneficial effects of being high in both (Kruglanski, Pierro, Mannetti, & Higgins, 2013). In this vein, individuals who are high in both locomotion and assessment have been shown to have higher GPAs (Kruglanski et al., 2000), greater chances of successfully completing an elite army training unit (Kruglanski et al., 2000), and better work performance (Pierro, Pica, Mauro, Kruglanski & Higgins, 2012). However, no previous studies have examined whether being high on both locomotion and assessment is always beneficial, or whether there are situations in which being high on both can actually be detrimental to performance. The aim of this paper is to address this gap in the literature by focusing on one potential moderator of the effects of regulatory mode on performance: task complexity. Cognitive task complexity is defined as any attentional, memory, reasoning, or other information processing demands that are imposed by the structure of a task (Robinson, 2001). We therefore conceptualize task complexity as a function of objective task characteristics (March & Simon, 1958; Schwab & Cummings, 1976). Some characteristics of more (vs. less) complex cognitive tasks include a greater amount of information to process (Campbell, 1988; Robinson, 2001; Schroder, Driver, & Streufert, 1967), the presence of multiple ways to attain the goal (Terborg & Miller, 1978), uncertainty of outcomes (March & Simon, 1958), more task dimensions requiring attention (Schroder et al., 1967), and higher rate of information change (Schroder et al., 1967). Given that locomotors have been shown to exhibit a speed-accuracy tradeoff (in which high locomotors are faster, but high assessors are more accurate; Kruglanski et al., 2000), we suggest that individuals who are high only on locomotion, only on assessment, or on both locomotion and assessment may differ in their aptitude for completing tasks that differ on some of the aforementioned characteristics. When a cognitive task involves less information processing, only a single way to achieve the goal, more certain outcomes, fewer dimensions that require attention, or a lower rate of information change (in other words, when the task is simple), it requires less thought to execute successfully (Campbell, 1988). As such, a simple task can be accomplished merely by starting it as soon as possible and moving continually toward its attainment. Individuals high on locomotion prefer to avoid procrastinating (Pierro et al., 2011) and tend to move continuously toward their goal, disliking obstacles or interruptions (Kruglanski, Pierro, & Higgins, 2015; Kruglanski et al., 2000). Thus, high locomotors should be particularly well-suited to performing simple tasks. On the other hand, individuals high on assessment tend to gather information extensively and compare many alternatives before coming to a decision (Kruglanski et al., 2000). Since simple tasks do not require such information gathering for optimal performance, high assessors are not necessarily well-suited to performing such tasks. Tasks which include a greater amount of information to process, the presence of several ways to attain the goal, uncertain outcomes, many task dimensions that require attention, or a higher rate of information change (in other words, cognitively complex tasks) require extensive information gathering and processing (Campbell, 1988). As such,
136
M. Chernikova et al. / Personality and Individual Differences 89 (2016) 134–142
complex tasks are best accomplished by critically evaluating the task and selecting a proper strategy before proceeding. Individuals high on assessment are therefore particularly well-suited to performing complex tasks, as they enjoy critical thought and careful evaluation (Higgins et al., 2003). In addition to this, locomotion also plays a role in complex task performance. Performance on all tasks, including complex ones, is improved when one starts the task without procrastinating and when one moves toward the goal continuously—both of which are characteristics of high locomotors (Kruglanski et al., 2000; Kruglanski et al., 2015; Pierro et al., 2011). Thus, individuals who are high on both locomotion and assessment should be the most suited for performing complex tasks. In accordance with the logic outlined above, we propose the following hypotheses: 1. When a task is simple, the combination of high locomotion and low assessment should lead to the best task performance. 2. When a task is complex, the combination of high locomotion and high assessment should lead to the best task performance. We tested these hypotheses in two empirical studies, one carried out in an organizational setting, the other in the lab. In the first study, we measured individuals' chronic locomotion and assessment tendencies, asked them to rate the complexity of their daily tasks at work, and obtained measures of their job performance from their supervisors. In the second study, we measured individuals' chronic locomotion and assessment tendencies, manipulated the task complexity of an inbox task they had to complete in the lab, and measured how well they performed on that task. In these two studies we focus specifically on information load, or the amount of information to be processed, as an operationalization of task complexity. However, we argue that the logic outlined here should apply to all types of complex tasks, not just those which require more (vs. less) information processing. The potential effects of other forms of task complexity on locomotors' and assessors' performance will be returned to in the General Discussion section. 2. Study 1 The goal of Study 1 was twofold: (1) to investigate our hypotheses about the interactive effect of locomotion and assessment on task performance in tasks of varied complexity, and (2) to examine these hypotheses in a real-world organizational setting. 2.1. Participants Ninety-four employees (49 women) from an Italian organization (a supermarket), divided into 9 work units, participated in this research on a voluntary basis. They did not receive any compensation in exchange for their participation. Employees' mean age was 35.35 years (SD = 10.52; Min = 19; Max = 66), their mean job tenure was 7.48 years (SD = 7.44; Min = 1; Max = 40), and the mean group size was 13.04 (SD = 4.47; Min = 3; Max = 18). Most of the participants (60.6%) had a high school diploma, 26.5% had a junior high school diploma, and 6.6% had a college degree. Participants were informed about the study and consented in writing to the use of their anonymized data. The study complied with the Declaration of Helsinki and was approved by the local ethics committee. 2.2. Procedure Participants filled out the Regulatory Mode Scale (Kruglanski et al., 2000) to measure their locomotion and assessment tendencies. This was followed by the Job Complexity Scale, a 4 item self-report scale designed to measure the amount of information processing needed at work. The major dependent variable was a measure of subordinates' performance as rated by their supervisors.
2.3. Materials 2.3.1. Locomotion and assessment Kruglanski et al. (2000) developed two scales designed to measure the individual difference tendencies of locomotion and assessment. Analyses using several samples from the United States and Italy revealed that the two scales were temporally stable and internally consistent. In addition, validation studies have shown that the two constructs are relatively independent and have strong predictive validity, as well as good convergent and divergent validity (Kruglanski et al., 2000). The Italian version of the locomotion and assessment scales (Kruglanski et al., 2000) constitutes two separate 12-item self-report measures designed to tap individual differences in these tendencies. Specifically, respondents rate the extent to which they agree with self-descriptive statements reflecting locomotion (e.g., “By the time I accomplish a task, I already have the next one in mind”) and assessment (e.g., “I spend a great deal of time taking inventory of my positive and negative characteristics”). Ratings are made on a 6-point Likert-type scale with the response alternatives anchored at the ends with 1 (strongly disagree) to 6 (strongly agree). We computed two composite scores (one for locomotion and one for assessment) by averaging across responses to each locomotion and assessment item, respectively. In this sample, the Cronbach's α was .74 for both the locomotion and assessment scales.
2.3.2. Job complexity Following prior research on the topic (e.g., Campbell, 1988), in this study we operationalized job complexity as the amount of information processing needed at work. To assess the degree to which their jobs required workers to attend to and process data or other information, we used the 4-item Information Processing scale. The Information Processing scale is a subscale of the larger Work Design Questionnaire, which was previously validated by Morgeson and Humphrey (2006). The items presented to participants were: (1) “The job requires me to monitor a great deal of information,” (2) “The job requires that I engage in a large amount of thinking,” (3) “The job requires me to keep track of more than one thing at a time,” and (4) “The job requires me to analyze a lot of information.” Items were answered on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). Cronbach's α for the four items in the present study was .63.
2.3.3. Performance Participants' performance was measured by supervisors' ratings of them on a 4 item instrument (adapted from Gould, 1979; Hobman, Jackson, Jimmieson, & Martin, 2011; Pierro, Pica, Kruglanski, & Higgins, 2014). Items included an assessment of quantity of performance, quality of performance, and an overall rating of performance. Specifically, we asked supervisors to appraise the performance of their employees (i.e., the research participants) over the last year1 by responding to the following items: (1) “In terms of percentage, to what extent were the employee's objectives reached during the last year?”, (2) “How do you evaluate the employee's overall performance in the last year?”, (3) “How do you evaluate the quality of the work he/she has done in the last year?”, and (4) “How do you evaluate the quantity or volume of the work he/she has done in the last year?” The first item was rated on a 10-point scale ranging from 1 (10%) to 10 (100%), and the last three items were rated on 10-point scales ranging from 1 (extremely negative) to 10 (extremely positive). We computed a composite performance score by averaging across the responses to each item (Cronbach's α = .97).
1
Note that, as mentioned above, the minimum of participants' job tenure was 1 year.
M. Chernikova et al. / Personality and Individual Differences 89 (2016) 134–142 Table 1 Descriptive statistics and variable intercorrelations for Study 1.
1. Locomotion 2. Assessment 3. Job complexity 4. Performance
M (SD)
1
2
3
4
4.89 (0.51) 3.10 (0.74) 3.89 (0.60) 7.78 (1.28)
– .03 −.13 .15
– .13 −.34⁎⁎
– −.10
–
N = 94. ⁎⁎ p b .01.
3. Results Table 1 contains a summary of descriptive statistics and zero-order correlations between the variables. The predictions regarding the effect on performance of interaction between regulatory mode orientations and job complexity were tested by means of a moderated multiple regression analysis. We used the PROCESS program for this analysis (Model 3, Hayes, 2013). In this moderated multiple regression analysis the main effects of locomotion, assessment, and job complexity, and all possible two-way interactions and the three-way interaction were entered. All variables were standardized and the interaction terms were based on these standardized scores. The main effects of gender, age, job tenure, and education were also entered as control variables. The overall regression was significant (R2 = .40, F(11, 82) = 5.05, p b .001); the results of this analysis are reported in Table 2. As can be seen in Table 2, the results show a main effect of locomotion on performance, with higher locomotion associated with greater performance. In line with our predictions, there is also a positive and significant effect of the interaction between locomotion and assessment on performance. Of greatest importance for the current hypothesis, the three-way interaction is significant. The positive sign of the three-way interaction suggests that, as predicted, the complementarity effect is stronger in high complexity jobs (Hypothesis 2). These findings are illustrated via the predicted mean values shown in Fig. 1a and b. Following the suggestion of Aiken and West (1991), the values were obtained using one standard deviation above and below the means of the relevant variables in the regression equation. The simple interaction analysis conducted to further understand the nature of the three-way interaction (Aiken & West, 1991) revealed that the interaction effect between locomotion and assessment, in predicting performance, was significant in the high complexity job (1 SD above the mean: b = .89, SE = .24, t = 3.68, p b .001), but not significant in the low complexity job (1 SD below the mean: b = −.25, SE = .24, t = −1.05, p = .298). Moreover, the simple slope analysis conducted to examine the twoway interaction (locomotion × assessment) only for those participants in the high complexity job (1 SD above the mean) revealed that the relationship between locomotion and performance was positive and significant for participants relatively high in assessment (1 SD above the mean: b = 1.30, SE = .37, t = 3.47, p b .001), whereas this relationship was negative and significant for participants relatively low in assessment (1 SD below the mean: b = −.47, SE = .19, t = − 2.19, p =
Table 2 Summary of moderated regression analysis in Study 1.
Locomotion Assessment Complexity Loc × Ass Loc × Complex Ass × Complex Loc × Ass × Complex Age Gender Education Tenure
b
Beta
SE
t
p
.31 −.22 .18 .32 .10 .20 .57 .00 −.30 −.09 .05
.24 −.17 .14 .20 .07 .19 .32 .01 .12 −.04 .33
.13 .12 .13 .15 .16 .11 .19 .01 .24 .21 .02
2.42 −1.82 1.38 2.09 .63 1.86 3.02 .09 −1.25 −.45 2.65
.017 .072 .171 .040 .529 .066 .003 .926 .214 .656 .001
137
.032). On the other hand, the relationship between assessment and performance was positive and significant for participants relatively high on locomotion (1 SD above the mean: b = .86, SE = .33, t = 2.56, p = .012), whereas this relationship was negative and significant for participants relatively low in locomotion (1 SD below the mean: b = −.91, SE = .26, t = − 3.52, p b .001). These results provide support for Hypothesis 2. Although the simple interaction was not significant in low complexity jobs, a simple slope analysis revealed that the simple slope is in the direction that we would predict, such that the relationship between assessment and performance was negative and significant for participants relatively high in locomotion (1 SD above the mean; b = −.68, SE = .25, t = −2.74, p = .007), whereas this relationship was not significant for participants relatively low in locomotion (1 SD below the mean; p = .59). On the other hand, the relationship between locomotion and performance was not significant for participants relatively high in assessment (1 SD above the mean; p = .91), whereas this relationship was positive and trending toward significance for participants relatively low in assessment (1 SD below the mean; b = .47, SE = .29, t = 1.64, p = .105). These results provide support for Hypothesis 1. We acknowledge that the structure of the data is nested (i.e., individual performance ratings are nested within units) and that this may raise the concern of non-independent data. Thus, to estimate how much of the total variance in performance ratings was betweenunit, we calculated the intraclass correlation coefficient (ICC). The result yielded a coefficient of .13, suggesting that only a small proportion of the variance in performance ratings was between unit. This ICC value close to zero implies the independence of the data and suggests that the ordinary least squares (OLS) approach used above can be sufficient. However, consistent with research (for a brief review, see Peugh, 2010) that has shown ICC values between .05 and .20 to be common in multilevel modeling applications in social science research, and following the recommendation of Hayes (2006), we also applied a multilevel modeling approach to the data, using maximum likelihood (ML) estimation. In the analysis, reported in Table 3, we entered as fixed all our level-one control variables (gender, age, education, and job tenure) and the main predictor variables (locomotion, assessment, job complexity, and the interaction between them). Locomotion, assessment, and job complexity were grand mean centered, and the interaction terms was based on these centered scores. We added the group size (level-two or group level variable) as a fixed control variable. Finally, only the intercept was a random effect, entered at the unit (group) level. Focusing our attention on the fixed effects, the results of this analysis revealed that group size was not related to performance. More important, consistent with above reported results, the analysis confirms: (1) a significant and positive main effect of locomotion on performance; (2) a significant and positive two-way interaction effect between locomotion and assessment; and, of greatest importance, (3) the significant and positive 3-way interaction effect between locomotion, assessment, and job complexity. Finally, the random effect was not significant (B = .03, SE = .05, Wald = .62, p = .533), providing no support for the notion that the intercepts varied between units. These findings effectively show that our conclusions here are not compromised by the potential dependency of observations. 3.1. Discussion The results of this study provide initial support for both of our hypotheses: when job tasks involved less complexity, those who were high on locomotion and low on assessment had the best performance (Hypothesis 1); on the other hand, when job tasks involved more complexity, individuals who were high on both locomotion and assessment had the highest performance (Hypothesis 2). It should be noted that since the results within low job complexity were only trending toward significance, this study alone does not allow us to draw firm conclusions regarding the influence of regulatory mode on performance under
138
M. Chernikova et al. / Personality and Individual Differences 89 (2016) 134–142
Fig. 1. a. Interactive effect of locomotion and assessment on job performance under low job complexity (Study 1). b. Interactive effect of locomotion and assessment on job performance under high job complexity (Study 1).
conditions of low job complexity. In addition, the first study was correlational in nature and therefore mute with respect to causality. Since work performance was measured retroactively (i.e., supervisors' ratings of participants' job performance over the last year), it is possible that participants' work performance may have influenced their regulatory mode levels. The next study was conducted to address these issues. 4. Study 2 The major goal of Study 2 was to extend the findings of the first study with the use of an experimental manipulation of task complexity. In addition, we used a different measure of performance: participants' number of correct responses in an inbox organization task. Table 3 Multilevel analyses.
Locomotion Assessment Complexity Loc × Ass Loc × Complex Ass × Complex Loc × Ass × Complex Age Gender Education Tenure Group size
b
SE
t
p
.27 −.22 .21 .30 .06 .21 .49 .00 −.22 −.10 .05 −.02
.12 .11 .12 .14 .15 .10 .18 .01 .23 .20 .02 .03
2.23 −1.93 1.70 2.11 .40 2.14 2.72 .18 −.92 −.49 2.78 −.67
.028 .056 .091 .037 .689 .035 .008 .858 .359 .626 .007 .507
4.1. Participants One hundred and three undergraduate participants (57 females) completed the study at Sapienza University of Rome in exchange for partial course credit. The average age of participants was 23.70 (SD = 2.83). There were no effects of gender or age, so they will not be discussed further. Participants were informed about the study and consented in writing to the use of their anonymized data. The study complied with the Declaration of Helsinki and was approved by the local ethics committee.
4.2. Procedure Participants filled out the Regulatory Mode Scale (Kruglanski et al., 2000) to measure their locomotion and assessment tendencies. They subsequently came into the lab and completed either a simple or complex inbox task. The major dependent variable was their performance on the inbox task.
4.3. Materials 4.3.1. Locomotion and assessment Participants completed the same regulatory mode questionnaire used in Study 1 in a mass testing session one month prior to coming into the lab. In this study, the Cronbach's α was .80 for the locomotion scale and .78 for the assessment scale. The two scales were not
M. Chernikova et al. / Personality and Individual Differences 89 (2016) 134–142
correlated in this sample, consistent with findings from previous studies (Kruglanski et al., 2000). 4.3.2. Task complexity Upon arriving in the lab, participants completed an inbox task that involved organizing and responding to various emails (previously used in Jimmieson & Terry, 1997; Jimmieson & Terry, 1999; Parker, Jimmieson, & Amiot, 2009; Parker, Jimmieson, & Amiot, 2013). All participants received a booklet containing instructions for the task and the pieces of information they needed to read and respond to (e.g., a team performance review and an organizational chart). Participants were instructed to assume the role of Alex Jennings, an Interim Retail Manager in a branch of the Borough Bank. They had to manage two team leaders and attempt to improve the branch's performance. In order to do this, they had to read a number of items (e.g., documents, emails, and memos) about human resources and financial issues. Participants had to write out the action they would take with regard to each of the items. They were told they should list as many actions as they considered necessary. Participants were randomly assigned to either the simple condition, in which they had to read and respond to six items (N = 52); or the complex condition, in which they had to read and respond to twelve items (N = 51). Independently of the task condition they were assigned to, all participants had a total of 90 min to complete the task. 4.3.3. Manipulation check After the inbox task, all participants filled out a 5 item manipulation check questionnaire of perceived task complexity (Cronbach's α = .80). The questionnaire contained one item measuring overall task complexity (“The task I have done is complex”) and 4 items measuring the amount of information required to process the task (e.g., “The task required me to manage a large amount of information”). After completing the manipulation check questionnaire, participants were provided with feedback on their performance on the inbox task; they were then thanked for their participation and dismissed. 4.4. Results We performed an ANOVA on the perceived complexity measure, which revealed that our manipulation of complexity was successful (F(1, 101) = 10.61, p = .002): participants in the simple condition reported lower levels of perceived complexity (M = 3.66, SD = 1.08) than those who were in the complex condition (M = 4.28, SD = .85). In addition, an ANOVA on the time measure (F(1, 101) = 28.09, p b .001) revealed that participants in the simple condition needed less time to complete the task (M = 39.04 min, SD = 20.20) than those in the complex condition (M = 62.37 min, SD = 24.33). This supports the notion that the simple task involved less information processing, as greater information processing is theorized to be more timeconsuming (Robinson, 2001). Performance on the inbox task was calculated by computing each participant's mean amount of correct actions per item. Whether the answers were marked as correct or not was established following a codebook written by the creators of the task (https://www.assessmentday. co.uk/answers-in-tray-assessmentday.pdf), which has been used in other published work on the inbox task (e.g., Jimmieson & Terry, 1997). For instance, one of the items participants had to address was an email from their subordinate, a sales executive, informing them that he was resigning due to a lack of developmental opportunities at the firm. Correct responses to this item included: (1) setting a meeting with the sales executive's direct supervisor to investigate any problems he may have had, (2) emailing the HR manager to gain an understanding of any issues that may have led to the resignation, and (3) assuring the staff that they will be offered more developmental opportunities from now on. Responses were scored for correctness by a trained coder who used the codebook, and who was blind to participants' locomotion and assessment scores.
139
Table 4 Descriptive statistics and variable intercorrelations for Study 2.
1. Locomotion 2. Assessment 3. Performance
M (SD)
1
2
3
4.42 (0.64) 3.84 (0.70) 1.60 (0.66)
– .07 .14
– −.02
–
N = 103.
Our predictions regarding the effect on task performance of the interaction between locomotion, assessment, and complexity condition were tested with a moderated multiple regression analysis. We used the PROCESS program for this analysis (Model 3; Hayes, 2013). Table 4 contains a summary of descriptive statistics and zero-order correlations between the variables. In the moderated multiple regression analysis the main effects of locomotion, assessment, and complexity condition, all possible two-way interactions, and the three-way interaction were entered. Following Aiken and West's (1991) recommendation, locomotion and assessment scores were mean centered and the interaction terms were based on these centered scores; task complexity condition was contrast coded as −1 (simple condition) and 1 (complex condition). The main effect of time was entered as a control variable, since participants differed in the amount of time they spent on the task in the simple and complex task conditions. Results of the moderated multiple regression analysis are reported in Table 5. The overall model was significant (R2 = .31, F(8, 94) = 5.40, p b .001). There was a main effect of condition, such that participants in the simple condition had better performance than those in the complex condition. There was also a main effect of time, such that the more time participants spent on the task, the better their performance. No other main effects were significant, nor were there any significant two-way interactions. Most importantly, as predicted, the main effects we observed were qualified by a significant three-way interaction (see Fig. 2a & b for interaction graphs). The simple interaction analysis conducted to further understand the nature of the three-way interaction (Aiken & West, 1991) revealed that the interaction effect between locomotion and assessment in predicting task performance was trending toward significance in both the simple condition (B = −.14, SE = .07, t(94) = − 1.96, p = .052) and the complex condition (B = .17, SE = .10, t(94) = 1.67, p = .097). The simple slope analyses revealed that when assessment was low in the simple task condition, high locomotors performed significantly better than low locomotors (B = .27, SE = .13, t(94) = 2.10, p = .038). When assessment was high in the simple condition, there were no significant differences between high and low locomotors (p = .76). These results support Hypothesis 1. When assessment was low in the complex task condition, there were no significant differences between high and low locomotors (p = .47). When assessment was high in the complex condition, high locomotors performed significantly better than low locomotors (B = .24, SE = .12, t(94) = 2.00, p = .048). These results support Hypothesis 2. When locomotion was low in the simple task condition, there were no significant differences between high and low assessors (p = .97). Table 5 Summary of moderated regression analysis in Study 2.
Locomotion Assessment Complexity cond. Loc × Ass Loc × Complex cond. Ass × Complex cond. Loc × Ass × Complex cond. Time
B
Beta
SE
t
p
.10 −.02 −.23 .00 −.02 .12 .16 .01
.14 −.03 −.34 .01 −.03 .18 .23 .51
.06 .06 .07 .06 .06 .06 .06 .00
1.65 −.37 −3.34 .15 −.39 1.97 2.51 5.21
.103 .714 .001 .884 .697 .052 .014 .000
140
M. Chernikova et al. / Personality and Individual Differences 89 (2016) 134–142
Fig. 2. a. Interactive effect of locomotion and assessment on performance in the simple task condition (Study 2). b. Interactive effect of locomotion and assessment on performance in the complex task condition (Study 2).
When locomotion was high in the simple condition, high assessors performed significantly worse than low assessors (B = −.30, SE = .12, t(94) = −2.56, p = .012). These results support Hypothesis 1. When locomotion was low in the complex task condition, there were no significant differences between high and low assessors (p = .616). When locomotion was high in the complex condition, there was a trend such that high assessors performed better than low assessors (B = .27, SE = .15, t(94) = 1.81, p = .072). These results support Hypothesis 2. 4.5. Discussion In summary, in the second study we found that high locomotion and low assessment led to better performance when a task was simple (supporting Hypothesis 1), whereas high locomotion and high assessment led to better performance when a task was complex (supporting Hypothesis 2). Importantly, although this study used a different task and a different measure of task performance, the results mirrored those from the first study. This suggests that our findings are not method-specific and can be extended to a variety of contexts. 5. General discussion Across two studies, we found that individuals high on locomotion and low on assessment performed better on simple tasks, whereas individuals high on both locomotion and assessment performed better on complex tasks. In the first study, when the work of employees in an Italian organization involved more information processing, those who were high on both locomotion and assessment had the best performance; on the other hand, when their work involved less information processing, those who were high on locomotion but low in assessment
had the highest performance. In the second study, university students who completed a simple inbox task in the lab performed best when they were high on locomotion and low on assessment. In contrast, students who performed a complex inbox task performed best when they were high on both locomotion and assessment. Taken together, the two studies suggest that high levels of both locomotion and assessment are necessary to successfully perform complex—but not simple—tasks. For simple tasks, high levels of locomotion alone are enough to ensure successful performance. To return to the story of Michael and Max, it seems that contrary to intuition, Max will not always be the best performer. Though the combination of careful thought and speedy action can often lead to high performance, there are some tasks which benefit most from a focus on swift action—and as such, individuals high on the trait of locomotion will have a significant advantage on such tasks. Thus, Max and Michael are each well-suited to performing tasks of different complexity levels, and would do well to seek out cognitive tasks that fit well with their regulatory mode orientation. The present studies extend the literature on regulatory mode in important ways. Most prior research on the interactive effects of locomotion and assessment has suggested that the high-high combination of locomotion and assessment is generally beneficial for both individual and team performance (e.g., Mauro et al., 2009; Pierro, Pica, Mauro, Kruglanski and Higgins, 2012; Pierro, Presaghi, Higgins, Klein and Kruglanski, 2012; see Kruglanski et al., 2013, for a review). However, contrary to prior research and theorizing on regulatory mode (e.g., Kruglanski et al., 2013), our studies show that being high on both locomotion and assessment may not always be advantageous for goal pursuit. Rather, differing levels of task complexity can determine whether an individual who is high only on locomotion, only on
M. Chernikova et al. / Personality and Individual Differences 89 (2016) 134–142
assessment, or high on both regulatory modes will have the highest performance on a task. Thus the current studies offer an important insight into the limits of the beneficial effects of a high-high regulatory mode combination. They also suggest that future theorizing on regulatory mode should explore the important question of which circumstances are most conducive to the performance of individuals who are high on assessment, high on locomotion, or high on both. This research can have intriguing implications for employee performance in organizations. An employer can evaluate the complexity of the tasks that a particular job requires and subsequently select employees based on their regulatory mode orientations. Individuals who are high on locomotion but low in assessment should be selected for tasks that are relatively simple, while those who are high on both locomotion and assessment should be selected for relatively complex tasks. The findings from these studies can also be extended to dyadic and group contexts: for instance, if two individuals are working together and are both high on locomotion, they should perform best on simple tasks. On the other hand, if one of the individuals is high on assessment and the other is high on locomotion, they should perform best on complex tasks. One limitation of these studies is that locomotion and assessment were measured and not manipulated in both studies; it would be important for this to be addressed in future research on the effects of locomotion and assessment complementarity. Another limitation is that both of our studies operationalize task complexity as workload, or the amount of information an individual has to process. While this operationalization of cognitive task complexity is supported by previous literature on complexity, there are also many other operationalizations of complexity which were not addressed in our studies (e.g., uncertainty of outcomes, multiple alternative paths to goal completion, and rate of information change, among others; Campbell, 1988). Given the lack of multiple operationalizations of complexity in the current studies, future research would do well to investigate the current hypotheses using different operationalizations of simple and complex tasks. A third limitation of this research is that both of our studies focused on cognitive task complexity as opposed to other types of complexity; as a result, we cannot speak to whether our results are generalizable to noncognitive tasks. Future studies could examine whether performance on tasks of varying physical complexity (e.g., assuming an easy vs. difficult yoga position) can also be influenced by individuals' regulatory mode. Other future directions for this line of research include investigating the interactive effect of regulatory mode complementarity and task complexity on outcomes besides performance. For instance, in addition to their higher performance on simple tasks, individuals who are high on locomotion and low on assessment may have higher levels of wellbeing and satisfaction when they work on simple tasks. On the other hand, individuals who are high on both locomotion and assessment may have higher levels of well-being and satisfaction when they work on complex tasks. In the organizational realm, when individuals experience a fit between their regulatory mode and the level of complexity their job requires, they may invest more effort in their job, exhibit fewer withdrawal behaviors (e.g., absenteeism, lateness), and cope better with organizational change. Future research can also examine other variables that moderate the effects of locomotion and assessment on performance: for example, high locomotors may perform better under time pressure than either high assessors or those who are high on both locomotion and assessment. Lastly, an anonymous reviewer made the insightful point that cognitive tasks can differ in information processing requirements in terms of whether those requirements pertain to planning (where planning requires deep thinking, but then execution does not) or execution (where planning is not feasible or helpful, and thus not complex, but execution requires complex information processing). Though the design of our current studies does not allow us to disentangle which of these aspects of task complexity was driving our effects, future research could separately manipulate increased
141
complexity during the planning stage, execution stage, or both stages of a task. It may be that high assessors are better at tasks that involve more complexity during the planning stage, high locomotors are better at tasks that involve more complexity at the execution stage, and individuals who are high on both locomotion and assessment are better at tasks which involve complexity during both the planning and execution stage. In conclusion, by elucidating the conditions under which various combinations of locomotion and assessment lead to higher task performance, the current studies offer a compelling addition to existing research on the effects of regulatory mode. These findings can have intriguing applications for organizations and other real-world contexts. They also provide support for the notion that regulatory mode is a useful and important social psychological construct, one that is relevant across many domains of research. References Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage. Amato, C., Pierro, A., Chirumbolo, A., & Pica, G. (2014). Regulatory modes and time management: How locomotors and assessors plan and perceive time. International Journal of Psychology, 49(3), 192–199. http://dx.doi.org/10.1002/ijop.12047. Avnet, T., & Higgins, E. T. (2003). Locomotion, assessment, and regulatory fit: Value transfer from “how” to “what”. Journal of Experimental Social Psychology, 39(5), 525–530. Campbell, D. J. (1988). Task complexity: A review and analysis. Academy of Management Review, 131(1), 40–52. Gould, S. (1979). Age, job complexity, satisfaction, and performance. Journal of Vocational Behavior, 14, 209–223. Hayes, A. F. (2006). A primer on multilevel modeling. Human Communication Research, 32(4), 385–410. Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York, NY: Guilford Press. Higgins, E. T., Kruglanski, A. W., & Pierro, A. (2003). Regulatory mode: Locomotion and assessment as distinct orientations. In M. P. Zanna (Ed.), Advances in experimental social psychology. 35. (pp. 293–344). New York: Academic Press. Hobman, E. V., Jackson, C. J., Jimmieson, N. L., & Martin, R. (2011). The effects of transformational leadership behaviours on follower outcomes: An identity-based analysis. European Journal of Work and Organizational Psychology, 20(4), 553–580. Jimmieson, N. L., & Terry, D. J. (1997). Responses to an in-basket activity: The role of work stress, behavioral control, and informational control. Journal of Occupational Health Psychology, 2, 72–83. Jimmieson, N. L., & Terry, D. J. (1999). The moderating role of task characteristics in determining responses to a stressful work simulation. Journal of Organizational Behavior, 20, 709–736. Kruglanski, A. W., Pierro, A., Mannetti, L., & Higgins, T. E. (2013). The distinct psychologies of “looking” and “leaping”: Assessment and locomotion as the springs of action. Social and Personality Psychology Compass, 7(2), 79–92. Kruglanski, A. W., Thompson, E. P., Higgins, E. T., Atash, M. N., Pierro, A., Shah, J. Y., & Spiegel, S. (2000). To do the right thing or to just do it!: Locomotion and assessment as distinct self-regulatory imperatives. Journal of Personality and Social Psychology, 79(5), 793–815. Kruglanski, A. W., Pierro, A., & Higgins, E. T. (2015). Experience of time by people on the go: A theory of the locomotion-temporality interface. Personality and Social Psychology Review, 1-18. Mannetti, L., Pierro, A., Higgins, E. T., & Kruglanski, A. W. (2012). Maintaining physical exercise: How locomotion mode moderates the full attitude–intention–behavior relation. Basic and Applied Social Psychology, 34(4), 295–303. March, J., & Simon, H. (1958). Organizations. New York: Wiley. Mauro, R., Pierro, A., Mannetti, L., Higgins, E. T., & Kruglanski, A. W. (2009). The perfect mix: Regulatory complementarity and the speed–accuracy balance in group performance. Psychological Science, 20(6), 681–685. Morgeson, F. P., & Humphrey, S. E. (2006). The Work Design Questionnaire (WDQ): Developing and validating a comprehensive measure for assessing job design and the nature of work. Journal of Applied Psychology, 91(6), 1321. Orehek, E., & Vazeou-Nieuwenhuis, A. (2013). Sequential and concurrent strategies of multiple goal pursuit. Review of General Psychology, 17(3), 339. http://dx.doi.org/10. 1037/a0032584. Orehek, E., Mauro, R., Kruglanski, A. W., & van der Bles, A. M. (2012). Prioritizing association strength versus value: The influence of self-regulatory modes on means evaluation in single goal and multigoal contexts. Journal of Personality and Social Psychology, 102(1), 22–31. http://dx.doi.org/10.1037/a0025881. Parker, S. L., Jimmieson, N. L., & Amiot, C. E. (2009). The stress-buffering effects of control on task satisfaction and perceived goal attainment: An experimental study of the moderating influence of desire for control. Applied Psychology: An International Review, 58, 622–652. Parker, S. L., Jimmieson, N. L., & Amiot, C. E. (2013). Self-determination, control, and reactions to changes in workload: A work simulation. Journal of Occupational Health Psychology, 18(2), 173. Peugh, J. L. (2010). A practical guide to multilevel modeling. Journal of School Psychology, 48(1), 85–112.
142
M. Chernikova et al. / Personality and Individual Differences 89 (2016) 134–142
Pierro, A., Pica, G., Mauro, R., Kruglanski, A. W., & Higgins, E. T. (2012a). How regulatory modes work together: Locomotion-assessment complementarity in work performance. Testing, Psychometrics, Methodology in Applied Psychology, 19, 247–262. Pierro, A., Presaghi, F., Higgins, E. T., Klein, K. M., & Kruglanski, A. W. (2012b). Frogs and ponds: A multilevel analysis of the regulatory mode complementarity hypothesis. Personality and Social Psychology Bulletin, 38(2), 269–279. Pierro, A., Giacomantonio, M., Pica, G., Kruglanski, A. W., & Higgins, E. T. (2011). On the psychology of time in action: Regulatory mode orientations and procrastination. Journal of Personality and Social Psychology, 101(6), 1317. Pierro, A., Pica, G., Kruglanski, A. W., & Higgins, T. E. (2014). Regulatory mode orientations and self-forgiveness. Unpublished manuscript, University of Rome “La Sapienza,” Rome, Italy.
Robinson, P. (2001). Task complexity, task difficulty, and task production: Exploring interactions in a componential framework. Applied Linguistics, 22(1), 27–57. Schroder, H., Driver, M., & Streufert, S. (1967). Human information processing. New York: Holt, Rinehart and Winston. Schwab, D., & Cummings, L. (1976). A theoretical analysis of the impact of task scope on employee performance. Academy of Management Review, 1, 23–35. Terborg, J., & Miller, H. (1978). Motivation, behavior and performance: A closer examination of goal-setting and monetary incentives. Journal of Applied Psychology, 63, 29–39.