Sustained attention failures are primarily due to sustained cognitive load not task monotony

Acta Psychologica 153 (2014) 87–94

Contents lists available at ScienceDirect

Acta Psychologica journal homepage: www.elsevier.com/ locate/actpsy

Sustained attention failures are primarily due to sustained cognitive load not task monotony James Head, William S. Helton ⁎ University of Canterbury, New Zealand

a r t i c l e

i n f o

Article history: Received 14 April 2014 Received in revised form 2 July 2014 Accepted 21 September 2014 Available online xxxx PsycINFO classification: 2221 2300 2330 2340 2346 4010

a b s t r a c t We conducted two studies using a modified sustained attention to response task (SART) to investigate the developmental process of SART performance and the role of cognitive load on performance when the speed-accuracy trade-off is controlled experimentally. In study 1, 23 participants completed the modified SART (target stimuli location was not predictable) and a subjective thought content questionnaire 4 times over the span of 4 weeks. As predicted, the influence of speed-accuracy trade-off was significantly mitigated on the modified SART by having target stimuli occur in unpredictable locations. In study 2, 21 of the 23 participants completed an abridged version of the modified SART with a verbal free-recall memory task. Participants performed significantly worse when completing the verbal memory task and SART concurrently. Overall, the results support a resource theory perspective with concern to errors being a result of limited mental resources and not simply mindlessness per se. © 2014 Published by Elsevier B.V.

Keywords: Attention Mindlessness Monotony Sustained attention Vigilance

1. Introduction Sustained attention or vigilance is the ability of an organism to maintain focused attention on a task and respond to the occurrence of rare critical targets (Warm, 1984). The early work of Norman Mackworth established that even highly motivated and trained operators have great difficulty maintaining optimal vigilance over time (Mackworth, 1948). In laboratory settings, vigilance is commonly measured using a Go/No-Go target detection task, whereby participants are required to respond to rare Go targets and withhold to numerous neutral No-Go stimuli. Generally, participants' performance becomes impaired with timeon-task. Lapses in vigilance are measured by errors of omission (non-responses to the target stimuli) and/or unusually slow responses to correct target stimuli (Davies & Tune, 1969). More recently researchers have begun to utilize other methodological approaches to measure lapses of sustained attention. For example, the Sustained Attention to Response Task (SART; Robertson, Manly,

⁎ Corresponding author at: Department of Psychology, University of Canterbury, Private Bag 4800, Christchurch, New Zealand. Tel.: +64 3 364 2998; fax: +64 3 3642181. E-mail address: [email protected] (W.S. Helton).

http://dx.doi.org/10.1016/j.actpsy.2014.09.007 0001-6918/© 2014 Published by Elsevier B.V.

Andrade, Baddeley, & Yiend, 1997) is often used in experimental and clinical environments to measure lapses in sustained attention (Docktree et al., 2004; Johnson et al., 2007; Manly, Robertson, Galloway, & Hawkins, 1999; Smallwood, Obsonsawin, & Heim, 2003). The SART differs from the traditional formatted vigilance task (TFT) mentioned above by inverting the relative proportion of Go and No-Go responses. Unlike the traditional vigilance task, participants in the SART are required to respond to numerous neutral stimuli and withhold their response to the rare critical targets (Robertson et al., 1997). Generally, simple numeric stimuli are used in the SART. For example, participants are tasked with withholding responses to a predefined numeric target (e.g., 3) and overtly responding to a larger digit set (e.g., 1, 2, 5, 7, 6, 7, 8, and 9) using a single button response. Lapses of attention in the SART are primarily measured by errors of commission (EC; inappropriately responding to the rare No-Go target). Errors of commission occur very quickly in the SART, within 4 min. The use of the SART has generated theoretical debates regarding the underlying cause of sustained attention lapses and has subsequently resulted in a debate regarding whether the SART is itself an appropriate measure of sustained attention (Carter, Russell, & Helton, 2013; Doneva & De Fockert, 2014; Grahn & Manly, 2012; Jonker, Seli, Cheyne, & Smilek, 2013; Staub,

88

J. Head, W.S. Helton / Acta Psychologica 153 (2014) 87–94

Doignon-Camus, Bacon, & Bonnefond, 2014; Staub, Doignon-Camus, Després, & Bonnefond, 2013). Although there has been decades of research on the topic of vigilance, there is still disagreement with concern to the cognitive mechanisms responsible for causing sustained attention lapses (Ariga & Lleras, 2011; Greene, Bellgrove, Gill, & Robertson, 2009; Helton & Warm, 2008; Rosenberg, Noonan, DeGutis, & Esterman, 2013). Two main theories have been put forth to explain lapses of sustained attention, the mindlessness, boredom, or monotony theories (underload theory) and conversely, mental fatigue or resource expenditure theories (overload theory). Proponents of mindlessness and boredom theories argue that vigilance tasks are cognitively undemanding and are monotonous. Moreover, it is this monotony that causes participants to withdraw their attention from the task (Robertson et al., 1997). In line with this perspective, most vigilance tasks employed in experimental research, dating from Mackworth's original clock task, have been objectively monotonous in nature (Hancock, 2013; Manly et al., 1999, 2004; Robertson et al., 1997). According to the proponents of the mindlessness-boredom theory, the vigilance decrement is the result of participants becoming bored or entering a mindless state as a result of the objectively monotonous nature of the vigilance task stimuli (Robertson et al., 1997). In other words, a lack of exogenous support of attention causes participants to disengage from the vigilance task and this results in errors. Therefore, according to the mindlessness-boredom theoretical account of sustained attention lapses, performance should significantly improve by reducing the monotony of the task by adding stimuli either to the task directly or to the background environment, thereby providing exogenous attention support into the task environment. Indeed, mindlessness-boredom theorists have argued that including sporadic sounds in addition to the SART has a refocusing effect (Manly et al., 2004). Manly et al. argue that additional stimuli occurring simultaneously with the SART reorients the participants' executive attention system back to the task which enables participants to appropriately withhold to the No-Go signals. Alternatively, resource theorists argue that maintaining vigilance is cognitively demanding and is thus resource dependent (Head & Helton, 2012, 2013a; Head, Russell, Dorahy, Neumann, & Helton, 2011; Helton, 2009). Consequently, participants' ability to maintain focused attention on a vigilance task is a function of the amount of mental resources available (Helton, 2009; Helton & Warm, 2008; Shaw, Satterfield, Ramirez, & Finomore, 2012; Shaw et al., 2013). Thus, as task time progresses mental resources are depleted more quickly than they are replenished, which is behaviourally manifested as the increasing lapses of attention (performance impairment). In addition, many resource theorists argue that the SART is not itself a measure of sustained attention, but rather a measure of participants attempting to resolve conflicting task requirements of responding both as fast and as accurately as possible (Peebles & Bothell, 2004). Therefore, resource theorists argue that the SART is a measure of impulsivity, motor control, or response strategy (Carter et al., 2013; Funke et al., 2013; Head & Helton, 2012, 2013a,b; Helton, 2009; Stevenson, Russell, & Helton, 2011). This issue has not been overlooked by mindlessness-boredom theorists and they have also expressed concerns regarding the ability to separate out sustained attention lapses from motor control (inhibition) errors in the SART (Seli, Cheyne, Barton, & Smilek, 2012; Seli, Cheyne, & Smilek, 2012; Seli, Jonker, Cheyne, & Smilek, 2013). Failures in motor control are likely due to the response requirement of the task (numerous quick button responses rarely interrupted), which generates a pre-potent ballistic motor routine that is difficult to inhibit (Head & Helton, 2012, 2013a,b; Helton & Russell, 2011; Manly et al., 1999; Robertson et al., 1997). Indeed, participants often report that they are fully aware of the No-Go targets; however, they are unable to physically stop their hand from responding to the target (Head & Helton, 2013a; see also Cheyne, Carriere, & Smilek, 2009). Disrupting the pre-potent ballistic routine by having participants strategically slow their responses can mitigate errors of commission on the SART (Peebles & Bothell, 2004). Many of the errors of commission in the SART are likely

due to a speed-accuracy trade-off (SATO) and response strategy. There is growing evidence for the motor control or response inhibition interpretation of the SART. As mentioned previously, when the SART is given, participants are instructed to respond as fast and accurately as possible. However, simply requesting participants to either emphasize speed or accuracy has a significant effect on performance (shifting the SATO). For example, manipulating the task instructions on the SART to emphasize participants to respond slower significantly decreases errors of commission (Seli, Cheyne, & Smilek, 2012). Additionally, controlling a participant's rate of response by using an auditory metronome to delay a response significantly improves commission error performance on the SART (Seli, Jonker, Solman, Cheyne, & Smilek, 2013), thus shifting the SATO. Some researchers have suggested trying to remove the SATO effect from the SART statistically (via correlational methods; see Seli, Cheyne, Barton, et al., 2012; Seli, Cheyne, & Smilek, 2012), and thus attempting to remove the contamination of the SATO on the SART as a measure of attention lapses Alternatively, another way to control the participant's rate of response, and thus reduce the effect of the SATO, is by reformatting the SART to force slower movement times. Head and Helton (2013a) manipulated stimuli location predictability and stimuli acquisition time by employing a modified point and click mouse SART. Stimuli uncertainty was manipulated by presenting a single numeric stimuli in a predictable or unpredictable location within one of four boxes presented in a cross pattern (see Fig. 1 for similar experimental paradigm). In the random location presentation, a single number stimulus was presented at random in one of the four boxes. In the clockwise condition, a single number stimulus was first presented in the top box followed by number stimuli occurring in adjacent boxes in a clockwise direction. Though number stimuli were randomly sampled (1–9), location of occurrence was entirely predictable. With concern to stimuli acquisition (i.e., how stimuli were selected), participants completed a manual selection and automatic selection condition. In the manual selection condition, participants were required to physically move the mouse cursor to the box containing the Go stimulus (e.g., 1–9 except for 8) and withhold responses to No-Go (8). Conversely, in the automatic selection condition, each box containing a number was automatically selected by the computer; however, a physical button response was still required if a Go stimulus was presented. Modifying how participants select the stimulus had a significant effect on SART performance. When participants were required to make a physical movement to the target (manual selection) errors of commission significantly decreased relative to the automatic selection condition (Head & Helton, 2013a). Head and Helton argue that manipulating the motor component of the SART affords more time for participants to withhold their response to the No-Go stimuli. Additionally, Head and Helton computed correlations between errors of commission and correct response times to Go stimuli in each of the 4 conditions (thus, examining the intersubject SATO). There were statistically significant negative correlations in each condition except in the random manual-select SART. More recently, Head and Helton (2013b), investigated the developmental process of SART performance using the clockwise (predictable) manual-select SART over 4 weeks (once a week). The clockwise manual-select SART enabled stimuli location to be predictable and thus facilitated greater movement speed-up with practice. Head and Helton predicted that as participants become more skilled at the task it would result in speeded response errors (increased errors of commission). However, if participants were aware of their performance, then they would be able to strategically shift their SATO. Additionally, participants also completed self-report measures of task-related and task-unrelated thoughts every session to determine whether conscious thoughts showed relationships with SART performance. As predicted, participants generally sped up on the task, resulting in increased commission errors. The participants were, however, strategic and the participants' speed and accuracy performance inversely oscillated over sessions. Correlational analysis showed robust negative correlations between response time and

J. Head, W.S. Helton / Acta Psychologica 153 (2014) 87–94

accuracy at both the intra-subject and inter-subject levels of analysis; however, commission errors failed to show any relationship with the self-report thought measures. Results from Head and Helton (2013b) indicate that when stimuli location is predictable and participants become more skilled at the task, speed and accuracy trade-offs occur. To further explore this, in study 1 we wanted to determine whether we could reduce the development of participants' speeded responses to Go stimuli by making the stimuli location unpredictable, thus using Head and Helton's (2013b) random location version of the manual-select SART. Therefore in the current investigation we examine the developmental process of the SART using the random manual-select paradigm previously used by Head and Helton (2013b) over repeated sessions. We suspect the random location manual selection SART should result in a task in which the SATO has a decreased influence on task performance. Similar to Head and Helton (2013b), we also wanted to explore whether self-reported thoughts have a relationship with SART performance over sessions in this modified version. In study 2, we test whether exogenous stimuli impacts the random location manual SART. As proposed by the mindlessness theorists attention attracting stimuli presented concurrently with the SART should improve performance due to reducing objective task monotony and by providing exogenous attention support. In other words, adding a secondary stimuli or secondary task to the main SART task should reduce objective stimuli monotony and thus bolster attention by providing exogenous (external to the person) support. We are not examining subjective monotony or how the person perceives the monotony of the task environment. The mindlessness-boredom perspective focuses on the causative role objective monotony of the task environment has on lapses of sustained attention not necessarily subjective monotony (see also Szalma, Schmidt, Teo, & Hancock, in press). Thus, in study 2, we explore whether having participants monitor and remember verbally presented words while concurrently completing the SART improves performance.

89

2. Study 1 methods 2.1. Participants Twenty-three people (16 females, 7 males) from a psychology course at the University of Canterbury participated for course credit. All participants had normal or corrected-to-normal vision. Participants' age range was between 21 and 45 years (M = 24 years; SD = 5.27).

2.1.1. Material and procedure Upon arrival participants were randomly assigned to an isolated individual cubical station. Once participants were assigned to their station, participants would use the same stations over the span of 4 sessions (one session per week). Participants were seated approximately 50 cm in front of a video display terminal (377 mm × 303 mm, 75 Hz refresh rate) that was mounted at eye level. Participant's head movements were not restricted in any way. Cell phones and wristwatches were surrendered at the start of each session each time. Participants completed a random manual-select SART which is visually similar to the clockwise manual selection SART used in previous studies (Head & Helton, 2013a,b); but differs in target predictability. Participants view four (35 mm × 35 mm) boxes arranged in a cross pattern. During the trial procedure, participants were first presented with the four empty boxes with a fixation marker (6 mm × 6 mm) in the center of the four boxes (see Fig. 1) for 200 ms. Participants were instructed to focus their attention on the fixation marker at the onset of each trial to refocus their attention to the center of the boxes each time. After the fixation period, participants gained control of a crosshair icon (10 mm × 10 mm) that appeared in the same location as the fixation marker. A single numeric target or neutral digit was presented in one of the four boxes at random for 250 ms followed by a 1000 ms answer period. During the 1000 ms inter-stimulus interval, participants were presented with four empty boxes. Thus, the total target onset interval was

+ +8

200 ms

250 ms

+

1000 ms

+8 +

+ 4 1450 ms Fig. 1. Random select SART within- and between-trial timing of events.

2

90

J. Head, W.S. Helton / Acta Psychologica 153 (2014) 87–94

1450 ms. During the numeric stimuli presentation, a single black 20-pt Courier New font digit was presented in the box. Participants were instructed to respond to digits 1–9 except for the target number 8. Target probability was 1 out 9 (p = .11) and neutral digits occurred (p = .89). Target and neutral number sampling was completely random; however, each digit (1–9) had equal probability of appearing in each of the four boxes. Participants were instructed to use a Saitek joystick to guide the cursor into a box containing a neutral digit and make their response. Participants were instructed to use their index finger of their dominate hand to press a trigger button on the Saitek joystick once the crosshair icon was registered inside of the box containing a neutral distracter. Once the crosshair icon was registered inside of the box the perimeter of the box would then become bolded to give the participant a visual feedback cue that the crosshair was in the box and ready for response (for similar procedure see Head & Helton, 2013a,b). The task duration was approximately 7 min with 288 trials. Upon completion of the SART, participants provided responses to a modified version of the Dundee Stress State Questionnaire (DSSQ; Matthews et al., 2002) each session on the computer. The DSSQ is a subjective-state questionnaire that measures fatigue, arousal, and stress. The questionnaire contains 10 scales in total; however, for the purpose of this study we only used Task-Related Thoughts (TRT) and Task-Unrelated Thoughts (TRT). Each subscale consisted of 8-items in which a fivepoint Likert scale was used for each subscale, 1 = “never”, 5 = “very often”. The total duration of the experiment was 10 min which included a 1 min practice trial to familiarize or refamiliarize the person with the task. Presentation of stimuli and response accuracy and timing were accomplished using E-prime 2.0 software (Psychology Software Tools, Pittsburgh, PA).

3. Results 3.1. Performance Participants who failed to complete all four sessions were omitted from the analysis (N = 4). For each individual for each session we calculated mean correct response times (correct response to Go stimuli), proportion of errors of commission (inappropriately responding to No-Go stimuli), and proportion of errors of omission (inappropriately withholding response to Go stimuli) as the dependent measures for participants who completed sessions 1–4. These data are displayed in Table 1. To investigate change over sessions we analyzed task performance using repeated measures analyses of variance. There was a significant sessions effect for errors of omission, F(3,54) = 5.49, p = .002, η2p = .23. As session numbers increased participants made less errors of omission. All other main effects failed to reach significance, p N .05. Lastly, we examined orthogonal cubic trends (see Head & Helton, 2013b), to determine whether response time and errors of commission inversely oscillate over sessions, both failed to reach significance, p N .05.

Table 1 Descriptive statistics (means; standard deviations in italics) for each task over the four weeks. Measure

Week 1

Week 2

Week 3

Week 4

Omission errors

0.24 0.19 0.06 0.06 732.01 107.70 2.38 0.76 1.35 0.41

0.16 0.14 0.04 0.07 708.89 115.41 2.02 0.55 1.63 0.59

0.14 0.13 0.05 0.06 693.21 122.10 1.91 0.58 1.77 0.68

0.10 0.09 0.03 0.07 691.11 112.40 1.74 0.51 1.68 0.65

Commission errors Response time Task-related thoughts Task-unrelated thoughts

3.2. Subjective-state over sessions Similar to the performance analyses, we omitted participants (N = 4) who failed to compete all 4 sessions. For each individual we calculated the average subjective-state scores for TUTs and TRTs. These data are displayed in Table 1. Due to the TUT and TRT scales being based on the same response scale, we analyzed both scales together using a 2 (response scale; TRT vs. TUT) × 4 (sessions) repeated measures analysis of variance. Participants reported having significantly greater TRT (M = 2.02; SE = .11) relative to TUT (M = 1.61; SE = .10), F(1,18) = 9.79, p = .006, η2p = .35. There was a significant scale by sessions interaction, F(3,54) = 8.09, p b .001, η2p = .31. The main effect for sessions failed to reach significance, p N .05. However, when we ran separate one-way analyses of variance, participants showed decreased TRT over sessions F(3,54) = 7.76, p b .001, η2p = .30, while TUT failed to reach significance. 3.3. Relationship correlates between SART performance and self-reports To examine the relationship between SART performance and selfreported mindfulness measures, we utilized a novel statistical correlations technique that enabled us to examine within- and betweensubject correlations. This statistical technique has been used in similar experimental paradigms (see Head & Helton, 2013a,b). First, Intersubject correlations were investigated by averaging each participant's performance metrics on the SART and their subjective-state responses over the 4 sessions. Correlations were then conducted between performance metrics and subjective-state responses. Removal of the intrasubject variance results in differences that can be attributed to more stable (or trait) individual differences. These correlations would indicate, for example, whether someone who generally tends to respond quickly also tends to make errors of commission. Secondly, intrasubject correlations were calculated by converting each participant's performance metric and subject-state response to a standardized z-score across the 4 sessions. All standardized z-scores were then combined across participants for the analysis (see Helton, Funke, & Knott, 2014). Removal of the inter-subject variance results in differences which can be attributed to within-subject sessional or state differences. Unlike inter-subject correlations, intra-subject correlations would indicate, for example, whether when someone responds quicker themselves, they are more likely to also make errors of commission. There were significant correlations between errors of omission and response time for both the inter- and intra-subject calculations. More importantly there was no sizeable correlation between response time and errors of commission, especially at the intra-subject level, at least in comparison to other versions of the SART (Head & Helton, 2013a,b). Additionally, there was no substantive relationship between any of the performance metrics and self-report responses at the inter- or intra-subject level (see Table 2 for intra- and inter-subject correlations). 4. Discussion In study 1 we examined the developmental process of the random manual-select SART to explore whether stimuli location uncertainty can disrupt the development of the pre-potent ballistic motor routine. Table 2 Correlations between variables (intra-subject above main diagonal; inter-subject below the main diagonal). EC Errors of commission (EC) Errors of omission (EO) Reaction time (RT) Task-related thoughts (TRT) Task-unrelated thoughts (TUT)

0.28 −0.41 0.00 −0.08

EO

RT

TRT

TUT

0.28

−0.07 0.47

0.21 0.34 0.07

0.11 −0.23 0.01 −0.26

0.52 0.13 0.19

−0.14 0.24

Note: Bold correlations would be significant, p b .05, for an N of 19.

0.24

J. Head, W.S. Helton / Acta Psychologica 153 (2014) 87–94

Participants repeatedly completed the modified SART 4 separate times whereby they guided a cursor and selected boxes containing number stimuli. Additionally, participants also completed a subjective-state thought questionnaire measuring task-related and task-unrelated thoughts each session. Generally, repeated practice with a task results in measurable improvements with task performance (Ericsson, Krampe, & Tesch-Rmer, 1993). As participant repeatedly completed the random manual-select SART they made significantly fewer errors of omission. Though there was not a significant change in response time over sessions, there was a trend of participants responding faster to Go stimuli. Additionally, participants' errors of commission did not increase over time, but rather remained relatively stable across sessions and failed to show an inverse oscillation relationship between response time and errors commission as found in Head and Helton (2013b) using a predictable location version of the task. Moreover, the non-significant correlation between response time and errors of commission suggest that the SATO had minimal effects on SART performance, especially at the intra-subject (within-subject) level of analysis. However, the significant positive correlation between errors of omission and response at both the inter- and intra-subject level may suggest that participants had difficulty selecting the box with the Go stimuli in the time allotment provided. This result does mirror more closely that of traditional vigilance tasks where more errors of omission are matched with slower responses (Head, Helton, Neumann, Russell, & Shears, 2011; Head, Russell, et al., 2011). Overall participants reported having a greater amount of TRTs relative to TUTs. Examining the self-reports (TRTs and TUTs) over sessions indicated that only TRTs decreased over sessions. The founders of the SART have argued that the task lulls participants into a mindlessness state which is behaviorally manifested as errors of commission (Robertson et al., 1997). Interestingly, there was no significant relationship between errors of commission and self-reported subjective states at the inter- or intra-subject level. Moreover, as stated above, the influence of the SATO was mitigated with the modified SART task. We suspect more of the observed errors using the modified SART may be due to actual failures of attention and not failures of response inhibition unlike the original unmodified SART (Grier et al., 2003). 5. Study 2 Relative to prior findings from Head and Helton (2013b), the influence of the SATO was significantly reduced in study 1. This was primarily corroborated by the non-significant correlations between errors of commission (EC) and response time and the relatively suppressed size of the correlations in comparison to the unmodified SART. However, regardless of the motor component of the task being controlled for, errors of commission to GO stimuli were still observed. Thus, in study 2 we wanted to further investigate the modified SART to elucidate the cognitive mechanisms responsible for the errors. In other words, when the motor component or response inhibition demands of the task are reduced by forcing slower responses, are errors on the SART the result of mindlessness due to objective task monotony, or limited information processing resources available to process information due to depletion or competition? From a mindlessness perspective, increasing exogenous support of attention by adding stimuli to the environment should eliminate performance impairments on the SART (Manly et al., 2004). Thus in study 2, we wanted to further investigate whether errors of commission can be mitigated by increasing exogenous support of attention by adding meaningful, attention-capturing auditory stimuli to the modified SART. According to the mindlessness perspective, adding more interesting or attention capturing stimuli to the SART should significantly increase performance on the SART by reorienting participants' attention back to the task and by making the task setting less objectively monotonous (see Manly et al., 2004). This does not, itself, guarantee any

91

change in subjective monotony, but the mindlessness theory is primarily framed in terms of objective environmental or task monotony (exogenous attention support). The issue of how objective monotony and subjective monotony relate demands further research, but is not necessary for testing the mindlessness theory. Conversely, based on the mental resource perspective, adding secondary auditory stimuli should compound errors on the SART only if there is a cognitive load from the stimuli due to increased demands on limited processing resources. Indeed, evidence for this has been provided by prior investigations by Helton and Russell (2011, 2013) who have used the abbreviated vigilance tasks coupled with either verbal or spatial working memory load conditions. The task required participants to remember and recall either a configuration of dots in space or a set of 4 letters while concurrently completing a traditional formatted vigilance task. Adding a secondary task (verbal or spatial) significantly impaired vigilance performance by increasing demands on working memory. In study 2 we wanted to utilize a paradigm similar to Helton and Russell (2011, 2013) to examine the two theoretical accounts of the lapses of sustained attention. However, we wanted to incorporate a task that was not visually interfering and relatively more complex than used by Helton and Russell (2011, 2013). Therefore, in study 2 we utilized an auditory word recall task developed by Green and Helton (2011). As will be discussed in more detail in the methods section below, the auditory word recall task involves participants listening to words and remembering them for later free recall. This is advantageous in that word stimuli are likely to capture participants' attention relative to simple stimuli and should bolster attention exogenously (see Head, Helton, et al., 2011; Head, Russell, et al., 2011). To control for the cognitive load of the additional auditory stimuli, participants also performed the modified SART with scrambled, unintelligible auditory stimuli. 6. Methods 6.1. Participants Twenty one (15 females, 6 males) from study 1 participated for course credit. All participants had normal or corrected-to-normal vision. No participants reported difficulties with hearing. Participants age ranged between 20 and 45 years (M = 24 years; SD = 5.55). 6.1.1. Procedure Upon arrival, participants surrendered cell phones and time keeping devices. Participants were assigned to the same cubicle station as study 1. All participants completed a single practice session whereby they received feedback on their performance on the manual-select SART. Upon completion of the practice session, participants were instructed to place a pair of Panasonic RP-HT161E-K headphones on. Participants completed three separate tasks involving listening and recalling words and completing a shortened version of the manual-select SART used in experiment 1. The tasks were counterbalanced for order. A shorter version of the manual-select SART was used to accommodate the 3 minute auditory stimuli clips. However, target and neutral stimuli probability were kept the same (i.e., 11% and 89%, respectively). Word stimuli used in the memory tasks consisted of 3 lists each containing 20 words. The words in each list were balanced for various psycholinguistic properties: number of syllables = 2, number of letters = 5–7, Kucera– Francis word frequency = 0–30, concreteness rating = 6–7, and meaningfulness = 6–8, (Paivio, Yuille, & Madigan, 1968). The auditory word stimuli were enunciated by a male native New Zealander (Green & Helton, 2011). Unintelligible versions of each word list were created by sound decomposition and random splicing (see Green & Helton, 2011). Auditory stimuli were presented to participants using headphones at 65 dB. In the dual-task condition, participants were instructed that they would have to listen to words and try to remember

J. Head, W.S. Helton / Acta Psychologica 153 (2014) 87–94

and recall them at a later time while completing the manual-select SART. In the auditory only condition, participants were given similar instructions with the exception that they only needed to stare at a blank black screen and listen to words for later recall. Upon completion of the dual-task and auditory condition, participants were instructed to write down each word they could remember using pencil and paper. Participants were given 90 s to free recall words. In the manual-select SART condition, participants completed the SART while listening to the unintelligible words. Participants were instructed that they did not have to free recall anything at the end of the SART. Correctly enunciated or scrambled words were presented in 8 second intervals throughout each task. A 14 second silence period preceded the first and last auditory stimulus in each list. To prepare participants for the first word stimulus, an auditory start warning signal was played which consisted of one high pitch tone followed immediately by four lower pitched tones. A warning signal was played at the end of the task to prompt the participant to begin the free recall. The second warning signal consisted of a single long pitch tone. Participants were exposed to the start and end warning tone prior to the start of the tasks. Total experiment duration was approximately 20 min.

7. Results 7.1. SART performance For each individual we calculated mean correct response times (correct response to Go stimuli), proportion of errors of commission (inappropriately responding to No-Go stimuli), and proportion of errors of omission (inappropriately withholding response to Go stimuli) as the dependent measures. To increase power, all participants were included in the analyses regardless of whether they completed all 4 preceding sessions, as an analysis did not reveal differences in performance on the SART as a function of session number completed. To determine the effect of memory load on SART performance, paired-sample t-tests were used between the memory load SART and no memory load SART conditions. Participants made significantly more errors of commission (M = .07; SD = .12) in the memory load condition relative to those who did not have to recall words (M = .02; SD = .04), t(20) = 2.09, p = .05, Mdifference = .05 95% CI [.00; .10]. Similarly, participants made significantly more errors of omission in the memory load SART (M = .16; SD = .13) relative to the no memory load SART (M = .13; SD = .11), t(20) = 2.12, p = .05, Mdifference = .03 95% CI [.00; .06]. Analysis of response time failed to show significant differences between load and no load, t(20) = 0.20, p = .84, Mdifference = 1.2 95% CI [− 10.8; 13.1] (see Fig. 2).

EC RT

10

700

6

690 680

4

660 650

0

No memory load SART

To gauge mental workload, we used the NASA Task Load Index (NASA-TLX, Hart & Staveland, 1988). The NASA-TLX is composed of 6 items measuring different facets of workload (e.g., mental, physical, temporal demand, performance, effort, and frustration). Participants indicate the level of workload for each item by responding to a scale ranging from 0 to 100. To calculate global workload, we averaged the 6 workload measures for each individual to create a single coefficient (see Warm, Dember, & Hancock, 1996, for similar procedure). A oneway repeated measures ANOVA was conducted on global workload for memory load SART, no memory load SART, and the words only recall condition. The overall effect was significant, F(2,40) = 17.17, p b .001, Ƞ2p = .46, see Fig. 3. In order to follow up on this significant results we compared the memory load SART with the no memory load SART, t(20) = 3.99, p = .001, Mdifference = 10.1, 95% CI [5.2; 16.7], the memory load SART with the word recall only task, t(20) = 6.57, p b .001, Mdifference = 16.7, 95% CI [11.4; 22.0], and the no memory load SART with the word recall only task, t(20) = 1.72, p = .101, Mdifference = 5.8, 95% CI [−1.2; 12.7]. 8. Discussion Though response time did not differ between conditions, participants concurrently monitoring words for later recall and completing the SART had significantly more errors of commission and omission relative to when participants completed the SART in isolation. Adding secondary meaningful or attention-capturing auditory stimuli failed to reorient participants' attention back to the task, but rather it compounded errors on the task. These results lend stronger support for a resource theory perspective of the casual mechanism responsible for errors on the modified SART. When the influence of the SATO was controlled for by slowing responses, the additional meaningful (attention capturing) auditory stimuli did not improve performance, but rather compounded errors on the SART. Moreover, participants self-reported that the dual-task was more demanding than completing the isolated tasks as measured by the

70

670 2

7.3. Subjective state

720 710

8

For each individual we calculated the number of words recalled by each participant for each condition. A paired-sample t-test was used to test whether word recall differed between dual-task (completing the SART while concurrently monitoring words) versus just monitoring for words without a secondary task. Participants recalled less words when completing the dual-task (M = 10.6; SD = 2.9) versus only monitoring for words without a secondary task, (M = 12.6; SD = 3.6), t(20) = 3.47, p = .002, Mdifference = 2.0 95% CI [0.8; 3.1].

730

Reacon me (ms)

Percent errors of commission

12

7.2. Free recall word performances

Memory load SART

Fig. 2. Reaction time and proportion of errors of commission on the memory load and no load condition; error bars depict standard error of the mean.

NASA-TLX global workload

92

60 50 40 30 20 10 0 Dual-task

Auditory only Condion

Scrambled words

Fig. 3. Subjective-state workload response on NASA-TLX; error bars depict standard error of the mean.

J. Head, W.S. Helton / Acta Psychologica 153 (2014) 87–94

93

global workload measure. All of these findings match expectations based on a resource theory account of attention lapses (errors).

whereas these comparative findings are very sensible given the resource theory perspective.

9. Overall discussion

References

In fairness to the mindlessness perspective of the SART, it could be argued that by modifying the SART to reduce the SATO characteristic of the original SART, we have fundamentally changed the task. The modified random location SART may no longer be monotonous enough to be an engine of mindlessness. The question, however, is can the SATO be detangled from mindlessness or is the SATO itself what is actually meant by the term mindlessness? Is mindlessness reducible to response strategy (see Peebles & Bothell, 2004)? As a thought experiment, imagine you could train an insect or arachnid to perform the SART and they made errors of commission, would these be attributed to mindlessness by mindlessness theorists? While resource theory has been criticized (Navon, 1984), it is an entirely materialist and naturalistic theory of attention and requires no references to mysterious conscious states that may never be verifiable by third-person interrogation (Uttal, 2001). If we were to broaden concerns regarding the meaning of mindlessness (or even boredom for that matter; see Hancock, 2013) to all vigilance tasks, we are confronted by similar conundrums. The major issue in traditional low-Go vigilance tasks is the decline in correct detections with time-on-task. From a mindlessness or boredom perspective this decline is due to the objective monotony of the task or the emergence of a subjective state presumably resulting from the objective monotony of the task (mindlessness or boredom). There is however a real puzzle. Recent research indicates dolphins are able to maintain constant levels of vigilance; dolphins do not suffer the vigilance decrement for at least five days (Ridgway et al., 2006, 2009). Rats, however, like people suffer the vigilance decrement (McGaughy & Sarter, 1995). Are rats and people subject to mindlessness, but dolphins are not? Are rats boredom prone, but dolphins are not? From a resource theory perspective the difference between rats and people and dolphins is completely explainable. Dolphins have a unique capacity to be able to alternate sleep and activation across their hemispheres in order for them to be able to swim continuously and breathe (Lyamin, Manger, Ridgway, Mukhametov, & Siegel, 2008). Dolphins essentially have two resource pools where one can rest and replenish and the other can be utilized. Rest is a proven mechanism to enable recovery of the decrement and a cornerstone of the resource theory of the decrement (Lim, Quevenco, & Kwok, 2013; Ross, Russell, & Helton, 2014). Terrestrial mammals are unfortunately not endowed with this capacity; hence our kind suffers the vigilance decrement. When we terrestrial mammals have maxed out our vigilance system there is no other system to rely on. Resource theory is at least a plausible explanation of this puzzle; mindlessness, whatever it means, may not be a plausible explanation. Conscious states occurring during vigilance should of course be explored, but psychologists should also be wary of the temptation to rely excessively on introspection and unverifiable self-reports. We also recommend human sustained attention researchers to familiarize themselves with the intriguing work being done by comparative researchers, which may cast some doubts on theories of sustained attention failures which appear to continuously emerge in the human literature under slightly different guises. This recommendation is not just directed to mindlessness theorists, but others who propose non-resource theories of the vigilance decrement (see Kurzban, Duckworth, Kable, & Myers, 2013). All of these theories are cast into doubt given the comparative literature, as they provide no rationale for the difference seen amongst species,

Ariga, A., & Lleras, A. (2011). Brief and rare mental “breaks” keep you focused: Deactivation and reactivation of task goals preempt vigilance decrements. Cognition, 118(3), 439–443. Carter, L., Russell, P. N., & Helton, W. S. (2013). Target predictability, sustained attention, and response inhibition. Brain and Cognition, 82(1), 35–42. Cheyne, J. A., Carriere, J. S. A., & Smilek, D. (2009). Absent minds and absent agents: Attention-lapse induced alienation of agency. Consciousness and Cognition, 18, 481–492. Davies, D. R., & Tune, G. S. (1969). Human vigilance performance. London: Staples Press. Docktree, P. M., Kelly, S. P., Roche, R. A., Hogan, M. J., Reilly, R. B., & Robertson, I. H. (2004). Behavioral and physiological impairments of sustained attention after traumatic brain injury. Cognitive Brain Research, 20, 403–414. Doneva, S. P., & De Fockert, J. W. (2014). More conservative go/no-go response criterion under high working memory load. Journal of Cognitive Psychology, 26(1), 110–117. Ericsson, K., Krampe, R. T., & Tesch-Rmer, C. (1993). The role of deliberate practice in the acquisition of expert performance. Psychological Review, 100(3), 363–406. Funke, G., Knott, B., Mancuso, V. F., Strang, A., Estepp, J., & Menke, L. (2013). Evaluations of subjective and EEG-based measures of mental workload. Communications in Computer and Information Science, 373(1), 412–416. Grahn, J. A., & Manly, T. (2012). Common neural recruitment across diverse sustained attention tasks. PloS One, 7(11), e49556. Green, A. L., & Helton, W. S. (2011). Dual-task performance during a climbing traverse. Experimental Brain Research, 215, 307–313. Greene, C. M., Bellgrove, M. A., Gill, M., & Robertson, I. H. (2009). Noradrenergic genotype predicts lapses in sustained attention. Neuropsychologia, 47(2), 591–594. Grier, R. A., Warm, J. S., Dember, W. N., Matthews, G., Galinsky, T. L., Szalma, J. L., et al. (2003). The vigilance decrement reflects limitations in effortful attention not mindlessness. Human Factors, 45, 349–359. Hancock, P. A. (2013). The problem of iatrogenically created psychological phenomena. American Psychologist, 68(2), 97–109. Hart, S. G., & Staveland, L. E. (1988). Development of NASA-TLX (task load index): Results of empirical and theoretical research. In P. A. Hancock, & N. Meshkati (Eds.), Human mental workload (pp. 139–183). Oxford, UK: North-Holland. Head, J., & Helton, W. S. (2012). Natural scene stimuli and lapses of sustained attention. Consciousness and Cognition, 21, 1617–1625. Head, J., & Helton, W. S. (2013a). Perceptual decoupling or motor decoupling? Consciousness and Cognition, 21, 913–919. Head, J., & Helton, W. S. (2013b). Practice does not make perfect in a modified sustained attention to response task. Experimental Brain Research, 232(2), 565–573. Head, J., Helton, W. S., Neumann, E., Russell, P. N., & Shears, C. (2011). Text-speak processing. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 55(1), 470–474. Head, J., Russell, P. N., Dorahy, M. J., Neumann, E., & Helton, W. S. (2011). Text-speak processing and the sustained attention to response task. Experimental Brain Research, 216(1), 103–111. Helton, W. S. (2009). Impulsive responding and the sustained attention to response task. Journal of Clinical and Experimental Neuropsychology, 31, 39–47. Helton, W. S., Funke, G. J., & Knott, B. A. (2014). Measuring workload in collaborative contexts: Trait versus state perspectives. Human Factors, 56(2), 322–332. Helton, W. S., & Russell, P. N. (2011). Working memory load and the vigilance decrement. Experimental Brain Research, 212, 429–437. Helton, W. S., & Russell, P. N. (2013). Visuospatial and verbal working memory load: Effects on visuospatial vigilance. Experimental Brain Research, 224(3), 429–436. Helton, W. S., & Warm, J. S. (2008). Signal salience and the mindlessness theory of vigilance. Acta Psychologica, 129, 18–25. Johnson, K. A., Robertson, I. H., Kelly, S. P., Silk, T. J., Barry, E., Dáibhis, A., et al. (2007). Dissociation of performance of children with ADHD and high functioning autism on a task of sustained attention. Neuropsychologia, 45, 2234–2245. Jonker, T. R., Seli, P., Cheyne, J. A., & Smilek, D. (2013). Performance reactivity in a continuous-performance task: Implications for understanding post-error behavior. Consciousness and Cognition, 22(4), 1468–1476. Kurzban, R., Duckworth, A., Kable, J. W., & Myers, J. (2013). An opportunity cost model of subjective effort and task performance. Behavioral and Brain Sciences, 36(06), 661–679. Lim, J., Quevenco, F. C., & Kwok, K. (2013). EEG alpha activity is associated with individual differences in post-break improvement. NeuroImage, 76, 81–89. Lyamin, O. I. L., Manger, P. R., Ridgway, S. H., Mukhametov, L. M., & Siegel, J. M. (2008). Cetacean sleep: An unusual form of mammalian sleep. Neuroscience and Behavioral Reviews, 32, 1451–1484. Mackworth, N. H. (1948). The breakdown of vigilance during prolonged visual search. Quarterly Journal of Experimental Psychology, 1, 6–21. Manly, T., Heutink, J., Davidson, B., Greenfield, E., Parr, A., & Ridgeway, V. (2004). An electronic knot in the handerkerchief: “Content free cueing” and the maintenance of attentive control. Neuropsychological Rehabilitation, 14, 89–116. Manly, T., Robertson, I. H., Galloway, M., & Hawkins, K. (1999). The absent mind: Further investigations of sustained attention to response. Neuropsychologia, 37, 661–670. Matthews, G., Campbell, S. E., Falconer, S., Joyner, L. A., Huggins, J., & Gilliand, K. (2002). Fundamental dimensions of subjective state in performance settings: task engagement, distress, and worry. Emotion, 2, 315–340.

94

J. Head, W.S. Helton / Acta Psychologica 153 (2014) 87–94

McGaughy, J., & Sarter, M. (1995). Behavioral vigilance in rats: task validation and effects of age, amphetamine, and benzodiazepine receptor ligands. Psychopharmacology, 117, 340–375. Navon, D. (1984). Resources—a theoretical soup stone? Psychological Review, 91(2), 216–234. Paivio, A., Yuille, J. C., & Madigan, S. A. (1968). Concreteness, imagery, and meaningfulness values for 925 nouns. Journal of Experimental Psychology, 76, 21–25. Peebles, D., & Bothell, D. (2004). Modelling performance in the sustained attention to response task. In Proceedings of the sixth international conference on cognitive modelling (pp. 231–236). Pittsburgh, PA: Carnegie Mellon University/University of Pittsburgh. Ridgway, S., Carder, D., Finneran, J., Keogh, M., Kamolnick, T., Todd, M., et al. (2006). Dolphin continuous auditory vigilance for five days. The Journal of Experimental Biology, 209, 3621–3628. Ridgway, S., Keogh, M., Carder, D., Finneran, J., Kamolnick, T., Todd, M., et al. (2009). Dolphins maintain cognitive performance during 72 to 120 hours of continuous auditory vigilance. The Journal of Experimental Biology, 212, 1519–1527. Robertson, I. H., Manly, T., Andrade, J., Baddeley, B. T., & Yiend, J. (1997). “Oops!”: performance correlates of everyday attentional failures in traumatic brain injured and normal subjects. Neuropsychologia, 35, 747–758. Rosenberg, M., Noonan, S., DeGutis, J., & Esterman, M. (2013). Sustaining visual attention in the face of distraction: a novel gradual-onset continuous performance task. Attention, Perception, & Psychophysics, 75(3), 426–439. Ross, H. A., Russell, P. N., & Helton, W. S. (2014). Effects of breaks and goal switches on the vigilance decrement. Experimental Brain Research, 1–9. Seli, P., Cheyne, J. A., Barton, K. R., & Smilek, D. (2012). Consistency of sustained attention across modalities: Comparing visual and auditory versions of the SART. Canadian Journal of Experimental Psychology, 66(1), 44–50. Seli, P., Cheyne, J. A., & Smilek, D. (2012). Attention failures versus misplaced diligence: Separating attention lapses from speed-accuracy trade-offs. Consciousness and Cognition, 21, 277–291. Seli, P., Jonker, T. R., Cheyne, J. A., & Smilek, D. (2013). Enhancing SART validity by statistically controlling speed-accuracy trade-offs. Frontiers in Psychology, 4, 1–8.

Seli, P., Jonker, T. R., Solman, G. J. F., Cheyne, J. A., & Smilek, D. (2013). A methodological note on evaluating performance in a sustained-attention-to-response task. Behavior Research Methods, 45, 355–363. Shaw, T. H., Funke, M. E., Dillard, M., Funke, G. J., Warm, J. S., & Parasuraman, R. (2013). Event-related cerebral hemodynamics reveal target-specific resource allocation for both “go” and “no-go” response-based vigilance tasks. Brain and Cognition, 82(3), 265–273. Shaw, T., Satterfield, R., Ramirez, R., & Finomore, V. (2012). A comparison of subjective and physiological workload assessment techniques during a 3-dimensional audio vigilance task. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 56(1), 1451–1455. Smallwood, J., Obsonsawin, M. C., & Heim, S. D. (2003). Task unrelated thought: The role of distributed processing. Consciousness and Cognition, 12, 452–484. Staub, B., Doignon-Camus, N., Bacon, E., & Bonnefond, A. (2014). Investigating sustained attention ability in the elderly by using two different approaches: Inhibiting ongoing behavior versus responding on rare occasions. Acta Psychologica, 146, 51–57. Staub, B., Doignon-Camus, N., Després, O., & Bonnefond, A. (2013). Sustained attention in the elderly: What do we know and what does it tell us about cognitive aging? Ageing Research Reviews, 12(2), 459–468. Stevenson, H., Russell, P. N., & Helton, W. S. (2011). Search asymmetry, sustained attention, and response inhibition. Brain and Cognition, 77, 215–222. Szalma, J. L., Schmidt, T. N., Teo, G. W. L., & Hancock, P. A. (2014s). Vigilance on the move: video game-based measurement of sustained attention. Ergonomics (in press). Uttal, W. R. (2001). The new phrenology: The limits of localizing cognitive processes in the brain. The MIT Press. Warm, J. S. (Ed.). (1984). Sustained attention in human performance. Chichester, UK: Wiley. Warm, J. S., Dember, W. N., & Hancock, P. A. (1996). Vigilance and workload in automated systems. In R. Parasurman, & M. Mouloua (Eds.), Automated and human performance: Theory and applications (pp. 183–200). Mahwah, NJ: Erlbaum.