Investigating factors that influence conventional distraction and tech-related distraction in math homework

Computers & Education 81 (2015) 304e314

Contents lists available at ScienceDirect

Computers & Education journal homepage: www.elsevier.com/locate/compedu

Investigating factors that influence conventional distraction and tech-related distraction in math homework Jianzhong Xu* Mississippi State University, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 2 August 2014 Received in revised form 25 October 2014 Accepted 25 October 2014 Available online 1 November 2014

This study examined high school students' distractions in math homework. A confirmatory factor analysis was conducted on the scores of six items regarding conventional and tech-related distractions. Data revealed that conventional and tech-related distractions were empirically distinguishable. Two multilevel models were performed, with each type of distractions as the dependent variable. Both types of distraction were negatively related to four student-level variables (homework effort, homework environment, learning-oriented reasons, and value belief). In addition, both were positively related to three student-level variables (time on videogame, peer-oriented reasons, and time on homework) and one class-level variable (time on homework). Meanwhile, tech-related distraction was positively associated with parent education, whereas conventional distraction was negatively associated with expectancy belief, affective attitude, and grade level. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Distraction Homework Math Self-regulation Volition

1. Introduction Although distraction frequently occurs during goal-oriented academic activities (Schmitz & Wiese, 2006; Wolters, 2011), it becomes more of a concern when students are required to complete academic tasks during after-school hours. This is particularly the case for homework, as it occurs in the middle of competing (often more appealing) activities (e.g., television, sports, and extracurricular activities), with less supervision, structure, and time constraints than in-class study (Cooper, Robinson, & Patall, 2006; Wolters, 2011; Xu, 2004). Adding to this concern is the fact that the increased prevalence of new media (e.g., IPad, laptop, smart phone, and tablet) presents a profound new challenge of shielding academic goal striving (i.e., completing homework) from unwanted distractions (Calderwood, Ackerman, & Conklin, 2014; Griffin, 2014; Richtel, 2010; Wallis, 2006). Yet, in spite of research showing that distraction has an adverse effect on task completion, knowledge acquisition, application, and academic performance (Hsu, Babeva, Feng, Hummer, & Davison, 2014; Jacobsen & Forste, 2011; Junco & Cotten, 2011), there have been few attempts to systematically investigate models of factors that influence homework distractions. The present investigation attempts to fill this gap in research on homework distraction. 2. Theoretical framework Students' engagement and persistence on goal-directed academic tasks (e.g., homework) often demand the use of volitional control to guard against distractions and aid task completion (Boekaerts & Corno, 2005; Corno, 2004; McCann & Turner, 2004). Volitional control is concerned mainly with issues of implementation that take place once a goal is set, to maintain the needed focus and effort to pursue that goal, and to protect the attention to follow through that goal in the face of various alluring temptations and competing personal striving (Boekaerts & Corno, 2005; Corno, 2004). Research and theorizing on volitional control suggest that volitional control may be affected by multiple variables. First, volition control is characterized by self-regulatory activities of persistent and purposive striving (e.g., structuring the workspace, bypassing barriers, and staying focused; Corno, 2004). Thus, volitional control to protect against homework distractions may be influenced by students' initiative in arranging their homework environment as well as their effort invested in completing homework. * Department of Leadership and Foundations, Mississippi State University, P.O. Box 6037, Mississippi State, MS 39762, USA. Tel.: þ1 662 325 2186; fax: þ1 662 325 0975. E-mail address: [email protected]. http://dx.doi.org/10.1016/j.compedu.2014.10.024 0360-1315/© 2014 Elsevier Ltd. All rights reserved.

J. Xu / Computers & Education 81 (2015) 304e314

305

In their study on the importance of volition in the learning process, Garcia, McCann, Turner, and Roska (1998) explicitly linked the expectancy-value theory (e.g., task value; Eccles, 1983) to intention formation, implementation, and protection. They posited that volitional control is affected by the enjoyment students experience while working on an academic task and utility value of the task. Informed by the expectancy-value theory, Warton (2001) stated the critical role of perceived task value in task completion, including (a) task interest (whether homework is viewed as interesting), (b) task importance and utility (the importance and usefulness of homework in fulfilling a number of goals), and (c) task cost (perceived opportunity costs resulting from doing homework, such as restricting time available for sports, extracurricular, and leisure activities). Therefore, volitional control to guard against homework distractions may be further affected by task value, including task interest, task importance and utility, and task cost. Furthermore, according to the expectancy-value theory (Eccles, 1983; Eccles & Wigfield, 2002), individuals are more likely to engage in an academic task if they believe they can be successful in performing the task (i.e., expectancy belief). Thus, individuals' use of volitional control to inhibit homework distractions may be also influenced by their expectancy for successfully completing homework. Taken together, this body of literature suggests that homework distractions may be affected by a range of factors (e.g., take value, value expectancy, homework environment, and homework efforts). Therefore, there is a need to include these variables in models of homework distractions. 3. Studies pertaining to homework distractions One line of literature finds that students continue to struggle with homework distractions well into the high school years and beyond (Benson, 1988; Cooper, Lindsay, & Nye, 2000; Pool, Koolstra, & van der Voort, 2003; Xu & Corno, 1998). These homework distractions have existed for a long time, including television; phone calls; pets wanting attention, barking, climbing on furniture, or being noisy; family visitors; siblings moving into and out of the study areas teasing or initiating questions; background yelling, conversations, and crying (e.g., small babies); disturbance from doorbells, washing machines, or vacuum cleaners; noise from stereos, radios, tape players, or musical instruments; feelings of tiredness and restlessness; and playing other things during a homework session (e.g., a cup or a toy). Over the last decade, the ever-present new media technology presents expanding webs of distraction to focusing and learning, with homework in particular (Calderwood et al., 2014; Dietz & Henrich, 2014; Foehr, 2006; Richtel, 2010; Sana, Weston, & Cepeda, 2013; Xu, 2008a). Based on media diary data from 694 students in grades 3e12, The Kaiser Family Foundation (Foehr, 2006) found that students were frequently doing something else (65% of the time) when their primary activity was doing homework on the computer. Interestingly but not surprisingly, 50% of the time doing homework on the computer as their primary activity was using another media (e.g., text messaging, surfing websites, using e-mail, and playing computer games and video games). Using time-diary and survey data from 1026 university students, Jacobsen and Forste (2011) investigated their electronic media use, including social-networking sites, cell phone, texting, and e-mail. The study found that about two-thirds of the students (62%) used some type of nonacademic electronic media while attending class, studying, or doing homework. It also found that electronic media use was negatively related to academic performance. Thus, many students are now “wired for distraction” (Richtel, 2010). New media technology may be more distracting, because of its motion and visual attraction (Griffin, 2014) and its seamless integration or intrusion of work, play, and social interaction (David, Kim, Brickman, Ran, & Curtis, 2014). In the case of homework, using technological devices may be more appealing, tempting, and distracting. Students' experience of homework are found to be predominantly negative. Compared with their experiences with classwork and other after-school activities, students tend to have low affect, motivation, and attention while completing homework (Leone & Richards, 1989; Shernoff & Vandell, 2007; Verma, Sharma, & Larson, 2002). Thus, using technological devices presents an easy outlet for coping with negative experience and boredom during homework completion (Calderwood et al., 2014), especially for “multitasking generation” (Wallis, 2006). An ongoing text exchange with a friend, for example, can be an appetitive activity that induces positive affect that offsets the boredom of homework (David et al., 2014). Another body of literature finds that several variables may influence homework distractions, including homework environment (Xu, 2010; Xu & Corno, 2003), student attitude (Calderwood et al., 2014; Cooper, Lindsay, Nye, & Greathouse,1998; Xu, 2008a), and student characteristics (David et al., 2014; Xu, 2010). For example, Xu (2010) examined multilevel models of homework distraction, based on survey data from 969 students in grade 8 (52 classes) and 831 students in grade 11 (45 classes) in US. At the individual level, homework distraction was negatively related to affective attitude (b ¼ .22, p < .01), homework environment (b ¼ .20, p < .01), academic achievement (b ¼ .08, p < .01), learning-oriented reasons (b ¼ .08, p < .05), homework interest (b ¼ .07, p < .05), and adult-oriented reasons (b ¼ .06, p < .05). Males, compared with females, reported statistically significant lower levels of homework distraction (b ¼ .30, p < .01). On the other hand, those students who spent more time on television (b ¼ .15, p < .01), extracurricular activities (b ¼ .10, p < .01), sports (b ¼ .06, p < .01), and paid jobs (b ¼ .05, p < .05) reported that they were more likely to be distracted while doing homework. In addition, homework distraction was positively related to peer-oriented reasons (b ¼ .10, p < .01). At the class level, students in grade 11 (compared with students in grade 8) were more likely to be distracted while doing homework (b ¼ .13, p < .05). Overall, these variables explained 25.2% of the variance in homework distraction at the individual level, 77.7% of the variance at the class level, and 28.3% of the total variance. In another related study, Calderwood et al. (2014) examined the homework distractions and media multitasking (e.g., non-homework related computer e-mail, Internet activities, and cell phone use) among 60 undergraduate students, based on data from a three-hour solitary homework session in a laboratory environment. The study linked homework distraction frequency and duration to negative and positive affect (i.e., the experience of positive or positive mood states), subjective fatigue (i.e., feelings of fatigue, sluggishness, stiffness or strain in neck or eyes), homework task motivation (i.e., motivation or effort to perform well on homework tasks), and selfefficacy (i.e., the confidence to concentrate on homework activities in the next hour). The results from zero-order correlations indicated homework distractions were negatively associated with homework task motivation and self-efficacy, but positively associated with negative affect.

306

J. Xu / Computers & Education 81 (2015) 304e314

4. The present investigation Although the study by Xu (2010) incorporated items relating to both conventional and tech-related distractions, it did not distinguish these two types of distractions. Nor did it examine whether these two types of distractions were empirically distinguishable. Although, the study by Calderwood et al. (2014) paid closer attention to tech-related distraction, it was conducted in a laboratory environment, thereby isolating from some conventional distractions (e.g., the physical presence of roommates). In addition, both studies used homework distraction as a general indicator across different school subjects. As several studies have investigated domain-specific homework models (e.g., self-regulation and effort; Hong, Peng, & Rowell, 2009; Trautwein, Ludtke, Schnyder, & Niggli, 2006), it would be beneficial to examine math homework distraction in the present investigation in that math is an important achievement domain with typically high demands for homework. For example, it has been shown that students tend to spend about onefifth to two-fifths of their homework time on math homework (e.g., Kitsantas, Cheema, & Ware, 2011; Pezdek, Berry, & Renno, 2002; Wong, Lam, Wong, Leung, & Mok, 2001). The present investigation aims to address these research gaps by focusing on the following two aspects. First, to determine whether conventional distraction and tech-related distraction are empirically distinct (one-factor model vs two-factor model), I conducted a confirmatory factor analysis to test the underlying factor structure of homework distractions. This line of research is highly significant, as (a) it directly addresses the critical question about whether new media technology emerges as a distinct type of distraction, and as (b) relevant literature implies that new media technology (compared with conventional distraction) is more enticing, tempting, and distracting (e.g., its characteristics of being manipulated, portable, interactive, networkable, and visually attractive; and its seamless integration or intrusion of work, play, and social interaction; Arnone, Small, Chauncey, & McKenna, 2011; David et al., 2014; Griffin, 2014; Matthews & Schrum, 2003). Second, I conducted two multilevel models, with each type of distraction as the dependent variable. These multilevel models included all the predictor variables in the previous study (Xu, 2010), with two exceptions: (a) there is no comparable variable relating to free lunch status in China (which may be viewed as a less concern, as homework distraction was not related to lunch status in the previous study; Xu, 2010); and (b) unlike their counterparts in US, secondary students in China do not hold paid jobs during afterschool hours. Other than these two exceptions, the multilevel models included five additional variables at the student level (homework effort, value belief, expectancy belief, time on videogame, and time on homework). Homework effort as a variable that may be negatively related to homework distraction has been alluded to by the literature on volitional control (e.g., Boekaerts & Corno, 2005) and previous empirical studies (e.g., Calderwood et al., 2014). Justification for including value belief and expectancy belief can be found in the expectancy-value theory (Eccles & Wigfield, 2002; Warton, 2001). Finally, as volitional control to guard homework distractions may be influenced by task cost (e.g., Garcia et al., 1998; Warton, 2001; perceived opportunity costs resulting from spending time on homework, such as restricting time available for videogame playing, TV, and extracurricular activities), it would be important to control time on homework and videogames in the present study (i.e., in addition to several variables included in the previous study such as time on sports, extracurricular, and TV). The present investigation further incorporated parent education and time on homework as two additional variables at the class level. The justification for including these two variables is that the social and academic contexts may influence how students approach homework (e.g., norms and expectations regarding homework; Corno & Mandinach, 2004) in that parent education and time on homework may have an effect on homework distraction above and beyond their effects at the student level. 5. Method 5.1. Participants and procedure The participants in the present investigation were 1799 high school students in China, including 915 students in grade 10 (23 classes) and 884 students in grade 11 (23 classes). The average class size was 39, ranging from 29 to 45. Particularly, of these students, 44.9% were male and 55.1% were female. Nearly all participants (97.3%) stated receiving math homework five or more days per week. In addition, the participants stated spending about 67 min doing math homework daily (SD ¼ 33); 64 min for students in grade 10 (SD ¼ 31) and 71 min for students in grade 11 (SD ¼ 34). These information concerning the amount of and frequency of math homework is similar to relevant findings from recent homework studies in China (OECD, 2010; Peng, Hong, Li, Wan, & Long, 2010). Before administrating the survey for this study, research assistants obtained standardized test scores from math teachers. The students were then assigned an identification number to ensure confidentiality and to link prior math achievement to the homework survey administered several months later. The homework survey was conducted in classroom settings independent of doing the actual homework. 5.2. Measures The participants were asked about educational level for father (or guardian) and mother (or guardian), from elementary school (scored 6 years) to graduate degree (scored 19 years). A composite variable for parent education was formed by averaging the educational levels for the parents. Students also indicated the extent to which they received family homework help, from never (scored 1) to routinely (scored 5). A number of multi-item scales were applied in the current investigation. Table 1 includes sample items for these scales, along with the reliability information from the present investigation. 5.2.1. Arranging the environment Five items assessed student initiative in selecting and structuring the homework environment (Xu, 2007, 2008b). This scale was developed, based on literature on self-regulation (Wolters, 2003; Zimmerman & Martinez-Pons, 1990). It ranged from finding a quiet and conducive workspace to locating relevant materials for math homework (a ¼ .72).

J. Xu / Computers & Education 81 (2015) 304e314

307

Table 1 Reliability estimates of multi-item scales. Scales

Sample items

a (CI)

HW environmenta

Find a quiet area Turn off the TV I do my best on my math HW I always try to finish my math HW I look forward to math HW I enjoy math HW My motivation to do math HW is ____d other after-school activities My attention while doing math HW is ___d other after-school activities Doing math HW gives you opportunities to work with classmates Doing math HW brings you approval from classmates Doing math HW brings you family approval Doing math HW brings you teacher approval Doing math HW helps you learn to work independently Doing math HW helps you understand what's going on in class Our math HW takes a lot of time and is of little use to mee There is no point in my doing math HWe I know exactly what I have to do at home to keep up in math lessons If I don't understand something in math, I know where to look it up How much of your math HW is checked by math teacher? How much of your math HW is collected by math teacher? Daydream during a math HW session Start conversations unrelated to what I'm doing Stop math HW to ply online games or videogames Stop math HW to send or receive text messages

.72 (.70e.74)

HW effortb HW interestc Affective attitude b

Peer-oriented reasons

Adult-oriented reasonsb Learning-oriented reasonsb Value belief

b

Expectancy belief

b

Teacher feedbackf Conventional Distractiona Tech-related Distraction

a

.80 (.78e.81) .94 (.94e.95) .82 (.81e.84) .67 (.64e.69) .83 (.82e.84) .90 (.90e.91) .83 (.82e.84) .80 (.79e.82) .69 (.67e.71) .77 (.75e.79) .74 (.72e.76)

Note. HW ¼ homework. a Rating: Never ¼ 1, Rarely ¼ 2, Sometimes ¼ 3, Often ¼ 4, Routinely ¼ 5. b Rating: Strongly disagree ¼ 1, Disagree ¼ 2, Agree ¼ 3, Strongly agree ¼ 4. c Rating: Strongly disagree ¼ 1, Disagree ¼ 2, Neither disagree nor agree ¼ 3, Agree ¼ 4, Strongly agree ¼ 5. d Rating: Much lower than ¼ 1, Lower than ¼ 2, About the same as ¼ 3, Higher than ¼ 4, Much higher than ¼ 5. e The item was reverse scored. f Rating: None ¼ 1, Some ¼ 2, About half ¼ 3, Most ¼ 4, All ¼ 5.

5.2.2. Homework effort Four items measured math homework effort, adapted from the work by Trautwein et al. (2006). These items focused on students' deliberate attempts to follow through their homework (e.g., doing best on their math homework; a ¼ .80). 5.2.3. Homework interest Five items measured students' interest in math homework (a ¼ .94), based on literature on interest (e.g., Denissen, Zarrett, & Eccles, 2007; Eccles & Wigfield, 2002) as well as homework interest (Cooper et al., 1998; Xu, 2008a). The scale assessed the extent to which students enjoyed doing math homework. 5.2.4. Affective attitude toward homework Four items measured students' favorability of math homework, as compared with their subjective experiences while engaging in other after-school activities, regarding their mood, attention, and motivation (Xu, 2008a; a ¼ .82). 5.2.5. Reasons for doing homework Three subscales measured reasons for doing homework (Xu, 2010). Three items measured peer-oriented reasons for math homework assignments (a ¼ .67), with respect to seeking approval from and working with their peers. Three items assessed adult-oriented reasons (a ¼ .83), with respect to seeking approval from their parents and math teachers. Nine items assessed learning-oriented reasons (a ¼ .90), with respect to forming good study habits and reinforcing math learning. 5.2.6. Value belief This scale included six items to measure students' perceived value of math homework, adapted from the work by Trautwein et al. (2006). These items focused on utility and cost of doing math homework (a ¼ .83). 5.2.7. Expectancy belief Adapted from the work by Trautwein et al. (2006), this scale included ten items to assessed students' expectancy belief regarding math homework (e.g., their confidence to complete math homework correctly; a ¼ .80). 5.2.8. Teacher feedback Based on related literature (e.g., Xu, 2008a), this scale included three items to measure how much math homework was monitored by teachers (a ¼ .69; the amount of math assignments being collected and checked). 5.2.9. Conventional distraction This scale taps into typical homework distractions that have existed for a long time (e.g., Pool et al., 2003; Schmitz & Wiese, 2006;Wober, 1992; Xu & Corno, 1998, 2003). It consisted of the following three items (a ¼ .77), including (a) daydream during a math homework session, (b) start conversations unrelated to what I'm doing, and (c) stop math homework to watch a favorite TV show.

308

J. Xu / Computers & Education 81 (2015) 304e314

5.2.10. Tech-related distraction (New media technology related distraction) The development of this scale is influenced by related literature on expanding webs of temptation and distraction relating to new media technology (e.g., Dietz & Henrich, 2014; Foehr, 2006; Richtel, 2010; Wallis, 2006; Xu, 2010). This scale consisted of the following three items (a ¼ .74), including (a) stop math homework to play online games or videogames, (b) stop math homework to send or receive email, and (c) stop math homework to send or receive text messages. The survey was translated and back-translated by several researchers who are proficient in both English and Chinese. Furthermore, Chinese teachers were asked to carefully examine the survey items to make sure that Chinese students clearly understand each item. 5.3. Statistical analyses First, a confirmatory factor analysis was conducted on the scores of six items relating to conventional distraction and tech-related distraction. The purpose is to examine whether conventional and tech-related distractions are empirically distinct, by comparing the difference in practical fit between a one-factor model (conventional and tech-related distractions are factorially indistinct) and a two-factor model (these two constructs are factorially distinct). Second, for each factor, I conducted multilevel analyses to take into consideration the nonindependence of observations by addressing the variability related to each level of nesting (i.e., at the individual and class level; Raudenbush & Bryk, 2002). To help interpret the regression coefficients, I standardized all continuous variables (M ¼ 0, SD ¼ 1) prior to conducting the multilevel analyses. Thus, the regression coefficients for the variables (except two dummy-coded variables e grade level and gender) were approximately comparable with the standardized regression coefficients in multiple regression analysis. All models implemented in this investigation were random-intercept models (i.e., the random parts of the slopes were not estimated; Raudenbush & Bryk, 2002) since we had no a priori hypotheses about if or how the predictive power of the level-1 predictor variables would differ across the classes. The multilevel models implemented in this investigation were described below. For the outcome of conventional distraction Yij, the level 1 of Model 1 for modeling students' conventional distraction can be described as below (using common notations in multilevel analysis, such as those in Raudenbush & Bryk, 2002):

      S þ rij Yij ¼ b0j þ b1j X1S þ b2j X2S þ …bkj X18 S represent eighteen student-level (indicated where Yij is the conventional distraction measure for student i nested under class j, and X1S to X18 by the superscript S) predictors. For instance, students' conventional distraction Yij could be related to gender (X1S ), and may be affected by other student-level variables (e.g., prior math achievement: X2S ). In this analysis, this representative Level 1 model included all eighteen student-level predictors described previously relevant to conventional distraction outcome Yij. The term rij represents un-modeled individual residual, and b0j represents the class intercept. In the model above, the model intercept (b0j) represent the class level conventional distraction. The variability among the classes in the model intercept (b0j) is likely influenced by class-level predictors, as shown in the Level 2 of Model 1 below:

        b0j ¼ g00 þ g01 Z1G þ g02 Z2G þ g03 Z3G þ g04 Z4G þ m0j where m0j is the un-modeled class residual, and class-level variables (i.e., grade level, parent education, teacher feedback, and time on math homework) are represented byZ1G, Z2G , Z3G , andZ4G (class level indicated by the superscript G). In this model, the effects of class level predictors on the intercept of Level 1 model are captured by the path coefficients g01 through g04. The effects (g01 through g04) on the Level 1 of the model's intercept will, in turn, translate into the effects on individual students' conventional distraction Yij. Model 2 is identical to Model 1, except that, in Model 2, tech-related distraction was replaced as the dependent variable. Full maximum likelihood was used in these two models. Missing values ranged from .06% to 3.34% (M ¼ 1.32%, SD ¼ .92%), and they were imputed using the expectation-maximization approach. 6. Results To obtain a first impression of the data, Table 2 shows the descriptive statistics and Pearson correlations of the study variables. Conventional distraction was significantly related to fifteen predictor variables, except five variables at the individual level (gender, parent education, family homework help, peer-oriented reasons, and time on sports) and two variables at the class level (grade level and teacher feedback). Tech-related distraction was significantly related to nineteen predictor variables, except three individual-level variables (gender, prior math achievement, and family help). Fisher's rz transformation revealed that the correlation of time on videogame and tech-related distraction was statistically higher than the correlation between time on videogame and conventional distraction, z ¼ 2.57, p < .05. It further revealed that the correlation of grade level and tech-related distraction was statistically higher than the correlation between grade level and conventional distraction, z ¼ 2.10, p < .05. On the other hand, the correlation of expectancy belief and conventional distraction was statistically higher than the correlation between expectancy belief and tech-related distraction, z ¼ 3.32, p < .001. 6.1. Confirmatory factor analysis As data from the present investigation are multilevel in structure, a two-level confirmatory factor analysis was conducted to examine whether conventional and tech-related distractions are empirically distinct. The fit of the following two models was compared using EQS (Bentler, 2006): (a) a one-factor model (conventional and tech-related distractions are factorially indistinct), and (b) a two-factor model (these two constructs are factorially distinct). As indicated in Table 3, the chi-square difference test between these two models was highly significant, c2 (Ddf ¼ 2) ¼ 351.929, p < .001. In addition, the difference in practical fit between the two models (i.e., DCFI ¼ .109) was

Table 2 Descriptive statistics and correlations. Variables

Gender (male: 1) Prior math achievement Parent education Family HW help HW environment HW effort HW interest Affective attitude Peer-oriented reasons Adult-oriented reasons Learning-oriented reasons Value belief Expectancy belief Time on sports Time on extracurricular Time on TV Time on videogame Time on HW Grade level (11th: 1) Parent education - class Teacher feedback - class Time on HW - class Conventional distraction Tech-related distraction

M

SD .45

1

2

3

4

5

6

7

8

9

10

11

12

13

14

70.48 13.15

16

17

18

19

20

21

22

23

.06y e

15.17

2.47 .03

.02

2.02

.93 .02

.09y

3.69

.86 .09y .12y .01

3.35

.55 .07y

.08y

.03

.05*

.24y e

2.87

.90 .05*

.06*

.01

.13y

.16y

2.64

.76 .04

.07y

.01

.09y

.09y

.26y

.65y e

2.64

.57 .02

.03

.01

.15y

.12y

.15y

.29y

2.34

.67 .01

.01

.03

.21y

.14y

.17y

.33y

.30y

.66y e

2.87

.54 .11y

.00

.02

.07y

.23y

.38y

.43y

.38y

.57y

.62y e

3.16

.50 .15y

.01

.01

.19y

.39y

.32y

.30y

.23y

.19y

.56y e

2.89

.48

.16y

.09y

.08y

.42y

.42y

.38y

.19y

.24y

.34y

.02

.05*

.08y

.02

.03

.01

.07y

.07y e

.03

.03

.05*

.03

.02

.09y

.03

.06*

e .13y e .11y e

.03 .05*

.30y e .25y e

41.94 42.66

.22y .07y .02

.04

.02

41.43 46.23

.06y .02

.04

.05* .02

21.29 39.86

.10y

.01

.02

.06* .14y .09y .06* .01

18.34 40.14

.17y

.01

.03

.06* .18y .18y .06y

67.08 32.57

.07y .02

.01

.50

.51

.04

.04

15.16

.50

.01

.10y

3.86

15

.50 e

.35 .05

66.34 19.43

.05*

.05*

.09y

.05*

.05* .04 .20y .04

.05* .05*

J. Xu / Computers & Education 81 (2015) 304e314

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

.02

.00

.31y e

.03

.12y .17y .05

.30y

.02

.03

.14y .23y .06*

.24y

.31y

.44y e

.01

.00

.04

.02

.01

.03

.03

.03

.00

.02

.01

.04

.02

.01

.03

.02

.01

.02

.06y .26y .23y .09y .12y .16y .17y .26y

.01

.06*

.04

.04

.00

.02

.02

.05*

.00

.05* .07y .15y .11y .07y .07y .11y .08y .07y .04

.05*

.11y

.09y

.40y e

.04

.07y

.10y

.13y

.14y

.03

.11y .07y .09y .18y .23y .17y .11y .15y .22y .19y .16y

2.30

.94

.01

.07y

.00

.01

.20y .37y .21y .23y .04

.11y .26y .30y .36y

1.88

.82

.03

.03

.05*

.03

.19y .39y .22y .19y .06y .12y .29y .35y .26y

.35y e

.01

e .09y e .33y .24y e .17y .07y .26y e

.06y

.10y

.59y

.15y

.55y .28y e

.02

.08y

.15y

.22y

.18y .01

.07y .03

.06*

.11y

.15y

.30y

.16y

.06y .07y .21y .54y

.06*

.19y e

Note. HW ¼ homework. N ¼ 1799. *p < .05. yp < .01.

309

310

J. Xu / Computers & Education 81 (2015) 304e314

Table 3 Model comparison. Model

c2

df

CFI

RMSEA

RMSEA 90% CI

SRMR

Model comparison

Dc2

Ddf

1. One-factor model 2. Two-factor model

425.300

18

.874

.159

.146e.172

.050

e

e

e

DCFI e

73.371

16

.983

.062

.048e.077

.020

2 vs 1

351.929***

2

.109

Notes: N ¼ 1799 from 46 classes. CFI ¼ comparative fit index; RMSEA ¼ root mean square error of approximation; CI ¼ confidence interval; SRMR ¼ standard root mean squared residual. ***p < .001.

substantial, far exceeding the cutoff point (i.e., DCFI ¼ .01; Cheung & Rnesvold, 2002). Thus, conventional and tech-related distractions as two theoretical distinct constructs are empirically distinguishable. In other words, homework distraction comprises two separate yet related factors: conventional distraction and tech-related distraction. Coefficient alphas (along with their 95% CIs) for scores on conventional and tech-related distractions were .77 (.75e.79) and .74 (.72e76), respectively. These reliability estimates can be considered as adequate or respectable (DeVellis, 1991; Henson, 2001). As research shows that distraction adversely influences a range of desirable outcomes (e.g., task completion, knowledge application, and academic performance; Junco & Cotten, 2011; Jacobsen & Forste, 2011; Hsu et al., 2014), the major objectives in the present investigation were to examine (a) whether conventional and tech-related distractions are empirically distinct, and (b) models of variables that influence conventional and tech-related distractions. On the other hand, there is a need to examine this linkage in the case of homework (i.e., whether conventional and tech-related distractions were related to homework performance). Thus, the students were asked two additional questions, adapted from NELS: 88 and other related studies (e.g., Cooper et al., 1998). They were asked, “How much of your assigned math homework do you usually complete?” Ratings include 1 (none), 2 (some), 3 (about half), 4 (most), and 5 (all). They were further asked, “How often do you come to class without your math homework?” Ratings include 1 (never), 2 (rarely), 3 (sometimes), 4 (often), and 5 (routinely). Results from Pearson correlation coefficients revealed that conventional and tech-related distractions were positively related to the frequency of coming to class without math homework (conventional: r ¼ .25, p < .001; techrelated: r ¼ .30, p < .001) and negatively associated with the amount of math homework completion (conventional: r ¼ .25, p < .001; tech-related: r ¼ .24, p < .001). Taken together, these correlations were consistent with theoretical expectation and related empirical evidences (e.g., higher levels of instant messaging were related to lower levels of schoolwork completion; Junco & Cotten, 2011), thereby providing additional empirical support to the convergent validity of the homework distraction scale. 6.2. Multilevel analyses I conducted two multilevel models, with each type of distraction as the dependent variable (conventional distraction and tech-related distraction). Each multilevel model included eighteen individual-level variables (gender, prior math achievement, parent education, family help, homework environment, homework effort, homework interest, affective attitude, adult-, peer-, and learning-oriented reasons, value belief, expectancy belief, and time on sports, extracurricular, TV, videogame, and homework) and four class-level variables (i.e., grade level, parent education, teacher feedback, and time on homework). 6.2.1. Conventional distraction Conventional distraction was the dependent variable in the first model. The results from fully unconditional model showed that 7.0% of the variance in conventional distraction located at the class level. Eighteen individual-level and four class-level variables were incorporated in Model 1, as the use of multilevel modeling to control for cluster effects is warranted when ICCs are as low as .02 (e.g., Kreft & de Leeuw, 1998; Von Secker, 2002). As indicated in Table 4, the multilevel results showed that ten individual-level variables had a statistically significant effect on conventional distraction. Conventional distraction was positively related to peer-oriented reasons (b ¼ .12, p < .01), time spent on videogame (b ¼ .10, p < .01), and time spent on homework (b ¼ .07, p < .05). Meanwhile, conventional distraction was negatively related to expectancy belief (b ¼ .19, p < .01), homework effort (b ¼ .18, p < .01), homework environment (b ¼ .10, p < .01), learning-oriented reasons (b ¼ .10, p < .05), affective attitude (b ¼ .09, p < .01), and value belief (b ¼ .07, p < .05). Compared with females, males were less susceptible to conventional distraction (b ¼ .09, p < .05). At the class level, time spent on homework had a positive effect on conventional distraction (b ¼ .16, p < .01). Meanwhile, students in grade 11 (compared with students in grade 10) were less likely to be sidetracked by convention distraction (b ¼ .09, p < .05). Taken together, this model (Model 1) explained 21.4% of the variance in conventional distraction at the individual level, 90.9% of the variance at the class level, and 26.3% of the total variance. 6.2.2. Tech-related distraction In Model 2, tech-related distraction was replaced as the dependent variable. The findings from fully unconditional model indicated that 8.4% of the variance in tech-related distraction located at the class level. Eighteen individual-level and four class-level variables were then incorporated in Model 2. The results revealed that nine student-level variables had a statistically significant effect on tech-related distraction. Tech-related distraction was positively associated with time spent on videogame (b ¼ .20, p < .01), peer-oriented reasons (b ¼ .09, p < .01), parent education (b ¼ .07, p < .01), and time spent on homework (b ¼ .07, p < .05). Meanwhile, tech-related distraction was negatively related to homework effort (b ¼ .22, p < .01), value belief (b ¼ .12, p < .01), learning-oriented reasons (b ¼ .08, p < .05), and homework environment (b ¼ .06, p < .05). Compared with females, males were less susceptible to tech-related distraction (b ¼ .13, p < .01). At the class level, tech-related distraction was positively related to time on homework (b ¼ .14, p < .05). Overall, this model (Model 2) explained 21.6% of the variance in tech-related distraction at the individual level, 91.0% of the variance at the class level, and 27.5% of the total variance.

J. Xu / Computers & Education 81 (2015) 304e314

311

Table 4 Multilevel modeling results for homework distraction. Model Predictor

Individual level Gender (female: 0, male: 1) Prior math achievement Parent education Family HW help HW environment HW effort HW interest Affective attitude Peer-oriented reasons Adult-oriented reasons Learning-oriented reasons Value belief Expectancy belief Time on sports Time on extracurricular Time on TV Time on videogame Time on HW Class level Grade level (10th: 0, 11th: 1) Parent education Teacher feedback Time on HW R2 individual level R2 class level R2 total Deviance statistics Number of estimated parameters

Model 1: Conventional

Model 2: Tech-related

b

SE

b

SE

.09* .02 .02 .02 .10y .18y .06 .09y .12y .00 .10* -. 07* .19y .02 .03 .04 .10y .07*

.04 .02 .03 .02 .03 .02 .04 .03 .03 .03 .04 .03 .03 .03 .03 .03 .03 .03

.13y .00 .07y .04 .06* .22y .02 .03 .09y .02 .08* .12y .05 .02 .03 .02 .20y .07*

.04 .02 .02 .02 .03 .04 .03 .04 .03 .03 .04 .03 .03 .02 .02 .03 .03 .03

.09* .29 .05 .16y .214 .909 .263 4555.124 25

.04 .16 .07 .06

.03 .33 .03 .14* .216 .910 .275 4520.225 25

.04 .19 .04 .06

Note. HW ¼ homework. N ¼ 1799 from 46 classes. b ¼ unstandardized regression coefficient. SE ¼ standard error of b. R2 ¼ amount of explained variance. *p < .05. yp < .01.

7. Discussion The first objective of the present investigation was to examine whether conventional and tech-related distractions are empirically distinct. Results revealed that the two-factor model (conventional and tech-related distractions are factorially distinct) had a significant better fit than the one factor model (these distractions are factorially indistinct). In other words, conventional and tech-related distractions could not be collapsed into one factor with significant loss of information. Thus, the present investigation provided empirical support for the construct validity of the distinction between conventional and tech-related distractions. In line with the proposition that tech-related distraction may be categorically different from conventional distraction (e.g., David et al., 2014; Griffin, 2014), the present study suggests that both conventional and tech-related distractions need to be taken into account to better understand and handle homework distractions. Thus, the present investigation takes another important step forward in research on homework distraction, as previous research does not differentiate these two different types of distractions. The second objective of the present investigation was to examine a range of variables that affect conventional and tech-related distractions. Both conventional and tech-related distractions were negatively related to four student-level variables (homework effort, homework environment, learning-oriented reasons, and value belief). In addition, both conventional and tech-related distractions were positively related to three student-level variables (time on videogames, time on homework, and peer-oriented reasons) and one class-level variable (time on homework). Compared with males, females were more susceptible to conventional and tech-related distractions. Techrelated distraction was positively associated with parent education at the student level. On the other hand, conventional distraction was negatively related to two student-level variables (expectancy belief and affective attitude) and one class-level variable (grade level). The findings that both conventional and tech-related distractions were negatively associated with homework effort and homework environment are in line with related literature regarding the importance of volitional control (e.g., Boekaerts & Corno, 2005; Corno, 2004) and the finding relating to the importance of homework environment in handling homework distraction with US students (Xu, 2010). Similarly, the findings that both conventional and tech-related distractions were negatively related to value belief and learning-oriented reasons are consistent with the expectancy-value theory regarding the role of task importance and utility in dealing with homework distraction (Eccles, 1983; Warton, 2001) as well as the finding regarding the role of learning-oriented reasons on homework distraction with US students (Xu, 2010). In addition, the finding that females were more likely to be sidetracked by conventional and tech-related distractions is consistent with previous finding with US students (Xu, 2010) and college students (David et al., 2014). Although these findings are not surprising, the present investigation contributes significantly to the empirical evidence on homework distraction, by differentiating conventional distraction from tech-related distraction, by examining domain-specific homework distraction (i.e., math homework instead of homework across different subjects) in China, and by explicitly linking homework distraction to homework effort and value belief in multilevel models. A similar statement can be made about peer-oriented reasons in that peer-oriented reasons was positively associated with homework distraction with a sample of US students (Xu, 2010); it was positively related to both conventional distraction and techrelated distraction with Chinese students while doing their math homework in the present investigation.

312

J. Xu / Computers & Education 81 (2015) 304e314

It is intriguing to note that whereas time on TV, sports, and extracurricular were positively associated with homework distraction with US students (Xu, 2010), they were not associated with both conventional distraction and tech-related distraction in the present investigation. On the other hand, time on videogame was negatively related to both conventional distraction and tech-related distraction with Chinese students in the present investigation. These results, when taken together, suggest that the influence of TV, sports, and extracurricular on homework distractions are less apparent for Chinese students than for US students. In contrast, thoughts and temptations for playing videogames while doing math homework pose a more formidable challenge for Chinese students. How do we interpret the findings that both conventional and tech-related distractions were positively associated with time on homework at the individual and class levels? Given that spending more time on homework implies that students are likely to be exposed with various homework distractions, it is intriguing to find that previous homework studies have not investigated the linkage between time on homework and homework distractions. Thus, the present investigation extends previous homework research, by providing empirical support to the linkage between time spent on homework and two dimensions of homework distraction (i.e., conventional and tech-related distractions). What is significant about these results is that they have been revealed in a large sample of high school students in China while controlling for other covariates (e.g., prior math achievement) in multilevel analyses. These results regarding time on homework provide empirical support to the proposition that “just assigning ‘more’ homework is a mechanical response to a set of complex issues” (Epstein & Van Voorhis, 2001, p. 181), in the context of homework distraction. Meanwhile, how do we explain the findings that grade level (10th vs. 11th) was negatively associated with conventional distraction (but not associated with tech-related distraction) for Chinese high school students in China? These findings are not consistent with the study with US students (Xu, 2010) that grade level (8th vs. 11th) was positively associated with homework distraction. It is possible that the effect of grade level on homework distraction is moderated by societal and cultural influences, in the sense that Chinese education system is highly competitive after nine-year compulsory education (Chen, 2008) and that Chinese parents are more likely than US counterparts to encourage children to do well in math and math homework (Cai, 2003). Such encouragement is more pronounced for 11th graders than for 10th graders, as their children inch closer to the National College Entrance Examination in China. Thus, it is not surprising that students in grade 11 (compared with students in grade 10) were less susceptible to conventional distraction. How do we interpret the result that students in grade 11 (compared with students in grade 10) is not more likely to avoid tech-related distraction as well? One possible explanation is that tech-related distraction poses a profound new challenge for students (e.g., more appealing and tempting due to its capacities for active participation and for satisfying the need for instant human interaction; Calderwood et al., 2014; David et al., 2014; Richtel, 2010) and for their parents (e.g., more difficult to monitor due to their lack of technical expertise for using new media technology; Davies, 2011). This explanation is supported by the findings from the present investigation in that expectancy belief and affective attitude toward homework were negatively related to conventional distraction, but unrelated to tech-related distraction. This explanation is further substantiated by the present finding that parent education was positively associated with tech-related distraction (but unrelated to conventional distraction).; those students whose parents with higher education may have more access to new media technology, which makes them more vulnerable to tech-related distraction. 8. Limitations and implications for research The present investigation had addressed an important gap in research on homework, by providing empirical support for the construct validity of the distinction between conventional distraction and tech-related distraction, and by examining multilevel models of factors that influence conventional and tech-related distractions in math homework with Chinese students. However, it has some limitations that need to be considered when interpreting it findings. For example, its findings were largely derived from a cross-sectional survey (except for standardized math test scores, which was assessed several months prior to the survey). In addition, although great care was taken to control all major factors known to influence distractions (as informed by theoretical framework and related empirical findings), other predictor variables may have influenced conventional or tech-related distractions in math homework had they been included. Regarding future research, it would be important to refine and validate homework distractions in other settings, and with students at different developmental stages (e.g., at the middle school or college level). It would also be interesting to examine whether tech-related distraction can be further differentiated. For example, playing videogames may involve (a) console games such as Xbox, Wii, and Nintendo, and (b) games on a computer, tablet, or smart phone. The former requires more of a commitment (e.g., turning on TV, console, inserting the game, and then playing), whereas the latter can be initiated almost instantly. In addition, there is a need to examine a broad range of factors such as those identified in the present investigation in different subject areas in other cross-cultural settings, as relevant findings from the present investigation suggest that cultural and societal differences may influence homework distractions. Furthermore, it would be beneficial to conduct longitudinal studies that follow cohorts of students over extended periods of time to investigate how they classify and handle different types of homework distractions over time. It would also be important to use multiple data sources (e.g., videotaped observations, stimulated recall interviews, and trace logs) to gain a deep understanding of dynamic processes underlying different types of homework distraction. It would be highly desirable to investigate causal hypotheses by experimentally manipulating students' homework effort and by examining the effect of such a manipulation on conventional and tech-related distractions, as well as homework completion, homework performance, and academic achievement. Finally, as the demands and challenges associated with homework continue to evolve over time (e.g., online homework or group homework; Xu, Du, & Fan, 2013), there is a need to conduct qualitative studies to better understand how students approach and handle different types of homework distractions (Davies, 2011). 9. Implications for practice With respect to homework practices, the finding that homework distraction consists of two separate yet related factors (i.e., conventional and tech-related distractions) suggests that we need to pay closer attention to both the differences and similarities between conventional and tech-related distractions. Relating to the similarities, the findings that conventional and tech-related distractions were negatively related to four student-level variables (learning-oriented reasons, value belief, homework effort, and homework environment) suggest that

J. Xu / Computers & Education 81 (2015) 304e314

313

teachers can play a significant role in countering these distractions. For example, teachers may help students cope with these distractions, by making homework more purposeful and useful, by working with families to arrange a conducive homework environment (Xu, 2010), and by valuing and promoting students' effort in the homework process. What can we make of the findings that conventional and tech-related distractions were positively related to time on homework, time on videogames, and peer-oriented reasons? First, although it seems commonsensical that teachers need to place emphasis on homework quality over quantity, it is important to state that this recommendation is substantiated by the present findings in the context of homework distraction; homework quality (i.e., value and usefulness) may guard against conventional and tech-related distractions, whereas homework quantity (time on homework both at the individual and class level) may contribute to conventional and tech-related distractions. Second, as the temptation for playing videogames poses a major distraction, as videogames become increasingly accessible to students while doing homework (e.g., a smart phone) and increasingly difficult for parents to control or monitor, students need to learn to take more responsibility in this area. In addition, students need to assume more responsibility in coping with peer-related distraction while doing homework, as both conventional and tech-related distractions were positively associated with peer-oriented reasons. Relating to the differences between conventional and tech-related distractions, what can we take away from the findings that conventional distraction was negatively related to expectancy belief, affective attitude, and grade level, whereas tech-related distraction was positively associated with parent education? As discussed in the previous section, these findings taken together suggest that new media technology emerges as a distinct, more irresistible source of distraction, in the sense that expectancy belief, affective attitude, and grade level served as a deterrent to conventional distraction, but not to tech-related distraction. This is further evident in the finding that techrelated distraction was positively related to parent education, implying that students whose parents with higher education may have more exposure to new media technology, and yet, these students or their better educated parents have not found a way to effectively to cope with associated tech-related distraction. Consequently, to address tech-related distraction, it would be especially important to pay close attention to learning-oriented reasons, value belief, homework environment, and homework effort, given the present findings that no other variables served as significant deterrent to tech-related distraction in particular. It would also be desirable to pay closer attention to students' perspectives about what teachers or families can do to better assist them to guard against tech-related distraction. This is more of a case for females, as they were more likely to be sidetracked by tech-related distraction (as well as conventional distraction). Learning from students' voices would allow teachers to design and develop more purposeful homework assignments (e.g., task value) and inform families to provide more relevant assistance with homework (e.g., arranging the homework environment). This, in turn, encourages students to take a more proactive role (e.g., homework effort) in handling tech-related distraction in particular (e.g. developing with their implicit theories about favorable conditions to minimize tech-related distraction). Finally, as peer-oriented reasons was positively related to tech-related distraction, as much of tech-related distraction involve peers (e.g., send and receive text messages while doing homework), it would be critically important to promote and develop a norm of peer support on how to cope with tech-related distraction, by encouraging students to discuss, formulate, and share their successful strategies with their peers about guarding against the rising tide of tech-related distraction.

References Arnone, M. P., Small, R. V., Chauncey, S. A., & McKenna, H. P. (2011). Curiosity, interest and engagement in technology-pervasive learning environments: a new research agenda. Educational Technology Research and Development, 59, 181e198. Benson, R. (1988). Helping pupils overcome homework distractions. Clearing House, 61, 370e372. Bentler, P. M. (2006). EQS structural equations program manual. Encino, CA: Multivariate Software. Boekaerts, M., & Corno, L. (2005). Self-regulation in the classroom: a perspective on assessment and intervention. Applied Psychology: An International Review, 54, 199e231. Cai, J. (2003). Investigating parental roles in students' learning of mathematics from a cross-national perspective. Mathematics Education Research Journal, 15(2), 87e106. Calderwood, C., Ackerman, P. L., & Conklin, E. M. (2014). What else do college students “do” while studying? An investigation of multitasking. Computers and Education, 75, 19e29. Chen, J. J.-L. (2008). Grade-level differences: relations of parental, teacher and peer support to academic engagement and achievement among Hong Kong students. School Psychology International, 29, 183e198. Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-of-fit indexes for testing measurement invariance. Structural Equation Modeling, 9, 233e255. Cooper, H., Lindsay, J. J., & Nye, B. (2000). Homework in the home: how student, family, and parenting-style differences relate to the homework process. Contemporary Educational Psychology, 25, 464e487. Cooper, H., Lindsay, J. J., Nye, B., & Greathouse, S. (1998). Relationships among attitudes about homework, amount of homework assigned and completed, and student achievement. Journal of Educational Psychology, 90, 70e83. Cooper, H., Robinson, J. C., & Patall, E. A. (2006). Does homework improve academic achievement? A synthesis of research, 1987-2003. Review of Educational Research, 76, 1e62. Corno, L. (2004). Introduction to the special issue work habits and work styles: volition in education. Teachers College Record, 106, 1669e1694. Corno, L., & Mandinach, E. B. (2004). What we have learned about student engagement in the past twenty years. In D. M. McInerney, & S. V. Etten (Eds.), Research on sociocultural influences on motivation and learning: Vol. 4. Big theories revisited (pp. 299e328). Greenwich, CT: Information Age. David, P., Kim, J., Brickman, J. S., Ran, W., & Curtis, C. M. (2014). Mobile phone distraction while studying. New Media and Society. http://dx.doi.org/10.1177/1461444814531692. Davies, C. (2011). Digitally strategic: how young people respond to parental views about the use of technology for learning in the home. Journal of Computer Assisted Learning, 27, 324e335. Denissen, J. J., Zarrett, N. R., & Eccles, J. S. (2007). I like to do it, I'm able, and I know I am: longitudinal couplings between domain specific achievement, self-concept, and interest. Child Development, 78, 430e447. DeVellis, R. F. (1991). Scale development. Newbury Park, NJ: Sage. Dietz, S., & Henrich, C. (2014). Texting as a distraction to learning in college students. Computers in Human Behavior, 36, 163e167. Eccles, J. S. (1983). Expectancies, values, and academic behaviors. In J. T. Spence (Ed.), Achievement and achievement motives (pp. 75e146). San Francisco, CA: Freeman. Eccles, J. S., & Wigfield, A. (2002). Motivational beliefs, values, and goals. Annual Review of Psychology, 53, 109e132. Epstein, J. L., & Van Voorhis, F. L. (2001). More than minutes: teachers' roles in designing homework. Educational Psychologist, 36, 181e193. Foehr, U. G. (2006). Media multitasking among American youth: Prevalence, predictors, and pairings. Menlo Park, CA: Kaiser Family Foundation. Garcia, T., McCann, E. J., Turner, J. E., & Roska, L. (1998). Modeling the mediating role of volition in the learning process. Contemporary Educational Psychology, 23, 392e418. Griffin, A. (2014, March 21e22). Technology distraction and the learning environment. In Proceedings of the Southern Association for Information Systems Conference, Macon, GA. Henson, R. K. (2001). Understanding internal consistency reliability estimates: a conceptual primer on coefficient alpha. Measurement and Evaluation in Counseling and Development, 34, 177e189. Hong, E., Peng, Y., & Rowell, L. L. (2009). Homework self-regulation: grade, gender, and achievement-level differences. Learning and Individual Differences, 19, 269e276. Hsu, K. J., Babeva, K. N., Feng, M. C., Hummer, J. F., & Davison, G. C. (2014). Experimentally induced distraction impact cognitive but not emotional processes in think-aloud cognitive assessment. Frontiers in Psychology, 5(474), 1e9.

314

J. Xu / Computers & Education 81 (2015) 304e314

Jacobsen, W. C., & Forste, R. (2011). The wired generation: academic and social outcomes of electronic media use among university students. Cyberpsychology, Behavior, and Social Networking, 14, 275e280. Junco, R., & Cotten, S. R. (2011). Perceived academic effects of instant messaging use. Computers & Education, 56, 370e378. Kitsantas, A., Cheema, J., & Ware, H. W. (2011). Mathematics achievement: the role of homework and self-efficacy beliefs. Journal of Advanced Academics, 22, 310e339. Kreft, I., & de Leeuw, J. (1998). Introducing multilevel modeling. London: Sage. Leone, C. M., & Richards, M. H. (1989). Classwork and homework in early adolescence: the ecology of achievement. Journal of Youth and Adolescence, 18, 531e548. Matthews, D., & Schrum, L. (2003). High-speed internet use and academic gratifications in the college residence. Internet and Higher Education, 6, 125e144. McCann, E. J., & Turner, J. E. (2004). Increasing student learning through volitional control. Teachers College Record, 106, 1695e1714. OECD. (2010). Mathematics teaching and learning strategies in PISA. Paris: OECD. Peng, Y., Hong, E., Li, X., Wan, M., & Long, Y. (2010). Homework problems: do students from rural and urban schools perceive differently. International Journal of Learning, 17(3), 81e95. Pezdek, K., Berry, T., & Renno, P. A. (2002). Children's mathematics achievement: the role of parents' perceptions and their involvement in homework. Journal of Educational Psychology, 94, 771e777. Pool, M. M., Koolstra, C. M., & van der Voort, T. H. A. (2003). The impact of background radio and television on high school students' homework performance. Journal of Communication, 53, 74e87. Raudenbush, S., & Bryk, A. S. (2002). Hierarchical linear models applications and data analysis (2nd ed.). Thousand Oaks, CA: Sage. Richtel, M. (2010, November 21). Growing up digital, wired for distraction. The New York Times, A1. Sana, F., Weston, T., & Cepeda, N. J. (2013). Laptop multitasking hinders classroom learning for both users and nearby peers. Computers and Education, 62, 24e32. Schmitz, B., & Wiese, B. S. (2006). New perspectives for the evaluation of training sessions in self-regulated learning: time-series analyses of diary data. Contemporary Educational Psychology, 31, 64e96. Shernoff, D. J., & Vandell, D. L. (2007). Engagement in after-school program activities: quality of experience from the perspective of participants. Journal of Youth Adolescence, 36, 891e903. Trautwein, U., Ludtke, O., Schnyder, I., & Niggli, A. (2006). Predicting homework effort: support for a domain-specific, multilevel homework model. Journal of Educational Psychology, 98, 438e456. Verma, S., Sharma, D., & Larson, R. (2002). School stress in India: effects on time and daily emotions. International Journal of Behavior Development, 26, 500e508. Von Secker, C. (2002). Effects of inquiry-based teacher practices on science excellence and equity. Journal of Educational Research, 95, 151e160. Wallis, C. (2006, March 27). The multitasking generation: they're e-mailing, IMing and downloading while writing the history essay. What is all that digital juggling doing to kids' brains and their family life? Time, 167, 48e55. Warton, P. M. (2001). The forgotten voices in homework: views of students. Educational Psychologist, 36, 155e165. Wober, J. M. (1992). Text in a texture of television: children's homework experience. Journal of Educational Television, 18, 23e34. Wolters, C. (2003). Regulation of motivation: evaluating an underemphasized aspect of self-regulated learning. Educational Psychologist, 38, 189e204. Wolters, C. (2011). Regulation of motivation: contextual and social aspects. Teachers College Record, 113, 265e283. Wong, N. Y., Lam, C. C., Wong, K. M. P., Leung, F. K. S., & Mok, I. A. C. (2001). Students' views of mathematics learning: a cross-sectional survey in Hong Kong. Education Journal, 29(2), 37e59. Xu, J. (2004). Family help and homework management in urban and rural secondary schools. Teachers College Record, 106, 1786e1803. Xu, J. (2007). Middle school homework management: more than just gender and family involvement. Educational Psychology, 27, 173e189. Xu, J. (2008a). Models of secondary students' interest in homework: a multilevel analysis. American Educational Research Journal, 45, 1180e1205. Xu, J. (2008b). Validation of scores on the homework management scale for high school students. Educational and Psychological Measurement, 68, 304e324. Xu, J. (2010). Predicting homework distraction at the secondary school level: a multilevel analysis. Teachers College Record, 112, 1937e1970. Xu, J., & Corno, L. (1998). Case studies of families doing third grade homework. Teachers College Record, 100, 402e436. Xu, J., & Corno, L. (2003). Family help and homework management reported by middle school students. Elementary School Journal, 103, 503e518. Xu, J., Du, J., & Fan, X. (2013). “Finding our time”: predicting students' time management in online collaborative group work. Computers and Education, 69, 139e147. Zimmerman, B. J., & Martinez-Pons, M. (1990). Student differences in self-regulated learning: relating grade, sex, and giftedness to self-efficacy and strategy use. Journal of Educational Psychology, 82, 51e59.