Metacognitive beliefs mediate the relationship between mind wandering and negative affect

Personality and Individual Differences 107 (2017) 78–87

Contents lists available at ScienceDirect

Personality and Individual Differences journal homepage: www.elsevier.com/locate/paid

Metacognitive beliefs mediate the relationship between mind wandering and negative affect Richard Carciofo a,b, Nan Song c, Feng Du a,⁎, Michelle M. Wang d, Kan Zhang a a

Key Laboratory of Behavioral Science, Institute of Psychology, Chinese Academy of Sciences, Beijing, China Language Centre, Xi'an Jiaotong-Liverpool University, Suzhou, China School of English for Specific Purposes, Beijing Foreign Studies University, Beijing, China d Department of Psychology, Wellesley College, Wellesley, MA, USA b c

a r t i c l e

i n f o

Article history: Received 26 July 2016 Received in revised form 8 November 2016 Accepted 13 November 2016 Available online xxxx Keywords: Mind wandering Daydreaming Metacognitive beliefs Negative affect Mindfulness Sleep quality Mediation

a b s t r a c t Two studies (Ns = 254 and 130, aged 18–28) aimed to investigate associations between mind wandering and metacognitive beliefs, and whether these beliefs are involved in the relationship between mind wandering and negative affect. Participants completed questionnaire measures of metacognitive beliefs, mind wandering, daydreaming, negative affect, mindfulness, and sleep quality. Study 2 also included the Sustained Attention to Response Task, with thought-probe assessment of task-unrelated thought (mind wandering/daydreaming). The frequency of mind wandering/daydreaming/task-unrelated thought was found to positively correlate with the metacognitive dimensions of less cognitive confidence, more endorsement of belief in the uncontrollability/danger of thoughts, and more endorsement of belief in the need to control thoughts. Multiple-mediator analysis was undertaken with three main models where either mind wandering, daydreaming frequency, or task-unrelated thought was the predictor for negative affect. Metacognitive beliefs, mindfulness and sleep quality were simultaneously entered as potential mediators. Results showed that metacognitive belief in the uncontrollability/danger of thoughts was a consistently significant mediator, while mindfulness and sleep quality were less consistent. Overall, the current research indicates that metacognitive beliefs are an important consideration in the study of mind wandering/daydreaming, and a possibly key factor in understanding the association with negative affect. © 2016 Elsevier Ltd. All rights reserved.

1. Introduction Mind wandering, or daydreaming, involves attention becoming focused on mentation unrelated to the external environment or to any ongoing task (Schooler et al., 2011; Singer, 1966). There are wide individual differences, but mind wandering/daydreaming seems to occur frequently: thought-sampling of participants engaged in daily activities has found mind wandering occurring in around 20–50% of samples (e.g., Killingsworth & Gilbert, 2010; McVay, Kane, & Kwapil, 2009; Song & Wang, 2012). Mind wandering/daydreaming can occur with meta-consciousness/self-awareness (involving explicit awareness of the ongoing conscious experience), but may also occur without metaconsciousness (Schooler et al., 2011; Smallwood & Schooler, 2006). Meta-consciousness is a core aspect of mindfulness. Although there remains some disagreement about how to define mindfulness (Grossman & Van Dam, 2011), a central aspect is “… being attentive to and aware of what is taking place in the present” (Brown & Ryan, 2003, p.822). Mindfulness negatively correlates with the frequency of ⁎ Corresponding author at: Key Laboratory of Behavioral Science, Institute of Psychology, 16 Lincui Road, Chaoyang District, Beijing, 100101, China. E-mail address: [email protected] (F. Du).

http://dx.doi.org/10.1016/j.paid.2016.11.033 0191-8869/© 2016 Elsevier Ltd. All rights reserved.

mind wandering/daydreaming (rs ranging − 0.24 to − 0.46: Carciofo, Du, Song, & Zhang, 2014a; Mrazek, Smallwood, & Schooler, 2012; Stawarczyk, Majerus, Van der Linden, & D'Argembeau 2012), and they have been seen as opposing concepts, at least in relation to attentional control (Mrazek et al., 2012). Mindfulness is associated with better sleep quality (Howell, Digdon, Buro, & Sheptycki, 2008). In contrast, the frequency of mind wandering/ daydreaming is related to difficulty in sleep initiation (Ottaviani & Couyoumdjian, 2013), and other aspects of poor sleep quality, including more reported sleep disturbances, lower ratings of subjective sleep quality, and more daytime dysfunction (Carciofo, Du, Song, & Zhang, 2014b). Also, while mindfulness is associated with positive affect and well-being (Brown & Ryan, 2003; Giluk, 2009), mind wandering and daydreaming frequency are associated with negative affect and depression (e.g., Giambra & Traynor, 1978; Killingsworth & Gilbert, 2010; Smallwood, Fitzgerald, Miles, & Phillips, 2009). For example, in an experience sampling study with N2000 participants, the experience of mind wandering was a significant predictor of later negative mood (Killingsworth & Gilbert, 2010). The attention of many recent studies has focused on this relationship between mind wandering/daydreaming and negative affect (e.g., Marchetti, Koster, & De Raedt, 2012; Mason, Brown, Mar, &

R. Carciofo et al. / Personality and Individual Differences 107 (2017) 78–87

Smallwood, 2013; McMillan, Kaufman, & Singer, 2013; Ottaviani & Couyoumdjian, 2013; Ottaviani, Shapiro, & Couyoumdjian, 2013; Stawarczyk, Majerus, & D'Argembeau, 2013), and it may involve a wide range of influences, both contextual and temporal (Smallwood & Andrews-Hanna, 2013). Stawarczyk et al. (2012) found that mindfulness and encoding style (internal versus external focus of attention) mediated the association between daydreaming frequency (predictor) and psychological distress (criterion). Furthermore, poor sleep quality has also been found to be a mediator between mind wandering/ daydreaming frequency and negative affect (Carciofo et al., 2014b). However, a further possible influence on this relationship between mind wandering/daydreaming and psychological distress/negative affect could be metacognition, which refers to “… knowledge and cognition about cognitive phenomena …” (Flavell, 1979, p.906), including the processes, knowledge, and beliefs involved in the regulation of thought (Wells & Cartwright-Hatton, 2004). Thus, while metacognition includes meta-consciousness, it also includes metacognitive knowledge/beliefs about cognitive functioning, such as regarding intra-individual and inter-individual differences, and how these are related to goal-setting and strategy use (Flavell, 1979). The metacognitive approach to psychological disorder (e.g., Wells & Matthews, 1996; Wells, 2007) argues that metacognitive beliefs can produce maladaptive self-regulation (coping or response styles), such as frequent/extended worry or rumination. These maladaptive metacognitive beliefs and response styles/strategies are important in the development and maintenance of many psychological disorders, including anxiety, depression, and obsessions (Cartwright-Hatton & Wells, 1997; Wells, 2007; Wells & Cartwright-Hatton, 2004), and may also be involved in insomnia (Harvey, Tang, & Browning, 2005; Waine, Broomfield, Banham, & Espie, 2009). To assess individual differences in aspects of metacognition associated with psychological distress/disorder, Cartwright-Hatton and Wells (1997) developed the MetaCognitions Questionnaire (MCQ), with items related to, for example, confidence in cognitive functions (such as memory), beliefs that worry might be a helpful strategy in some situations, and beliefs that worry might be uncontrollable or dangerous. The success of the metacognitive approach developed by Wells and colleagues suggests that metacognitive beliefs are an important consideration for understanding many types of psychological distress. Although maladaptive daydreaming can occur when excessive fantasising limits social interaction and/or otherwise impairs daily functioning (Somer, 2002), mind wandering/daydreaming is not typically indicative of clinical disorder (Klinger, Henning, & Janssen, 2009; Singer, 1966). However, metacognitive beliefs might be involved in the widely reported relationship between mind wandering/ daydreaming and the experience of negative affect. Thus, the current research aimed to investigate how metacognitive beliefs, as identified in the metacognitive approach to psychological disorder (Wells & Matthews, 1996; Wells, 2007), are related to mind wandering/ daydreaming. In addition, it was investigated whether metacognitive beliefs mediate the relationship between mind wandering/ daydreaming frequency and negative affect. 2. Method 2.1. Materials Metacognitive beliefs were assessed with the 30-item version of the MetaCognitions Questionnaire (MCQ-30; Wells & Cartwright-Hatton, 2004). This has the same five-factor structure as the 65-item MCQ, and shows the same positive correlations with measures of anxiety, worry and obsessive symptoms. The MCQ-30 has six items for each of the following dimensions of metacognition: MCQ1 - Cognitive Confidence (e.g., “I have a poor memory”); MCQ2 - Positive Beliefs about Worry (e.g., “Worrying helps me cope”); MCQ3 - Cognitive Self-consciousness (e.g., “I monitor my thoughts”); MCQ4 - Negative Beliefs

79

about Uncontrollable Thoughts and associated Danger (e.g., “When I start worrying I cannot stop”); MCQ5 - the Need to Control Thoughts and Negative Beliefs about the Consequences of Thoughts (e.g., “I should be in control of my thoughts all of the time”). The order of the 30 items was randomised on the final questionnaire. Items are rated on a scale of: (1) do not agree; (2) agree slightly; (3) agree moderately; (4) agree very much. Higher scores for each dimension represent more maladaptive metacognitions (range = 6–24 for each dimension). The current Chinese version of the MCQ-30 was developed by back-translation: a native Chinese-speaker translated the original English scale, and another native Chinese-speaker backtranslated it; a native English-speaker checked the back-translation and discrepancies were resolved with the translators. The construct validity was assessed with Confirmatory Factor Analysis (CFA); test-retest (7–8 week interval) was also undertaken with a sub-sample from Study 1 (N = 114). Mind wandering and daydreaming were assessed with the following scales from the Imaginal Processes Inventory (Singer & Antrobus, 1972; Chinese versions: Carciofo et al., 2014a, 2014b): the Daydreaming Frequency scale (DF; e.g., “I lose myself in active daydreaming”), the Mind Wandering scale (MW; e.g., “I am the kind of person whose thoughts often wander”), and the Problem-Solving Daydreams scale (e.g., “My daydreams offer me useful clues to tricky situations I face”). Each scale has 12 items (6 reverse-scored on the MW scale; 3 reversescored for Problem-Solving Daydreams), each scored on 5-point Likert scales, giving scores ranging 12–60, with higher scores indicating more daydreaming frequency/mind wandering/problem-solving daydreams. While the DF and MW scales are associated with negative affect and depression, the Problem-Solving Daydreams scale has not shown such correlations (Carciofo et al., 2014b; Giambra & Traynor, 1978), so it was investigated whether this form of daydreaming is also differentially associated with metacognitive beliefs. Mindfulness was assessed with the 12-item Mindful Attention Awareness Scale-Lapses Only (MLO; e.g., “I find myself doing things without paying attention”). This is a shortened version of the 15-item Mindful Attention Awareness Scale (MAAS; Brown & Ryan, 2003), which omits one item related to driving (and so less relevant to students) and two items not related to lapses (Carriere, Cheyne, & Smilek, 2008; Chinese version: Carciofo et al., 2014a). Each item is scored on a 6-point Likert scale; total scores range 12–72, with higher scores indicating more frequent mindful states. The 12-item MLO scale correlates strongly with the 15-item MAAS (r = 0.961; Mrazek et al., 2012). Sleep quality was assessed with the Pittsburgh Sleep Quality Index (PSQI; Buysse, Reynolds, Monk, Berman, & Kupfer, 1989; Chinese version: Liu, Tang, Hu, et al., 1996), with components of: Subjective Sleep Quality, Sleep Latency, Sleep Duration, Sleep Efficiency, Sleep Disturbances, Use of Medication, and Daytime Dysfunction. Scale items are used to calculate a score (ranging 0–3) for each component, with higher scores indicating poorer quality sleep. Summing the seven components gives a global score (ranging 0–21). Participants completed the PSQI for their sleep over the preceding month. Affect was assessed with the Positive and Negative Affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988; Chinese version: Huang, Yang, & Ji, 2003); 10 items assess positive affect, with higher scores indicating more energy and ‘pleasurable engagement’, and 10 items assess negative affect, with higher scores indicating more general distress and ‘unpleasurable engagement’. Negative affect has shown moderate/strong positive correlations with measures of general psychological distress, depression and anxiety (Crawford & Henry, 2004; Watson et al., 1988). Each PANAS item is scored on a 5-point Likert scale, giving a range of 10–50 for each subscale. In Study 1, participants were instructed to complete the PANAS according to how they had felt over the preceding 3–4 weeks. In Study 2, participants were instructed to complete the PANAS according to how they felt “now”/at this moment.

80

R. Carciofo et al. / Personality and Individual Differences 107 (2017) 78–87

2.2. Missing data For each scale (including MCQ-30 and PANAS subscales), a single missing item was replaced by the participant's mean score for the other items. If a scale had been incorrectly completed (e.g., multiple answers for an item), or if there were two/more omissions, then the questionnaire was excluded. The PSQI components require some calculations based on item responses; consequently, missing or ambiguous data were dealt with according to a set of rules (available from the authors). Cases were omitted if they had missing scores for one/more of the PSQI components. 2.3. Participants and procedure Study 1 involved 254 undergraduate students (57 male, 197 female) from two Beijing universities, with a mean age of 18.77 (SD = 0.702; range = 18–21); male mean = 18.95 (SD = 0.718); female mean = 18.72 (SD = 0.691); t = 2.161, p = 0.032. A sub-group of these participants (N = 121) were part of the sample reported in Carciofo et al. (2014b, Study 3). Participants completed the MCQ-30, DF, MW, MLO, PSQI, and PANAS scales (and some others unrelated to the current aims, and not reported here), during class breaks. For most participants, the questionnaires were completed in two or three sessions, over approximately 5–8 weeks. A sub-group (N = 40) completed the MCQ30 approximately seven months after completing the other scales. The sequence of the scales within and between sessions was varied, to provide counter-balancing to some extent. Participation was voluntary and unpaid. Study 2 involved a separate sample of 130 participants (60 male, 70 female), with a mean age of 22.18 (SD = 1.955; range = 18–28); male mean = 22.92 (SD = 1.759); female mean = 21.54 (SD = 1.901); t = 4.251, p b 0.0005. Participants were mostly university students, recruited online or from other research studies, who completed an experiment involving several tasks, with aims unrelated to the current study. After signing the consent form, participants completed the first administration of the PANAS. They then completed five experimental tasks including the SART (see below; others were unrelated to the current aims, and are not reported here), and then completed the PANAS (second administration), DF, MLO, PSQI, MCQ-30, and MW scales (plus some others not reported here). They were then debriefed and remunerated for their participation. Analysis of the PANAS data used each participant's mean score from the two administrations. The Sustained Attention to Response Task (SART; Robertson, Manly, Andrade, Baddeley, & Yiend, 1997) required participants to press the space bar in response to the digits 1–9, except for withholding response to the target (number 3). Each digit was presented 28 times (target probability = 11%) for a total of 252 trials, organised into 18 blocks of between 8 and 26 digits/trials. Block order was randomised, with pseudo-random presentation of targets within each block. A fixation cross was shown (on a CRT screen) for 800 milliseconds at the beginning of each block, and then each digit was presented for 300 ms, followed by a blank (black) screen for 1200 ms. An on-screen thought-probe was presented at the end of each block, asking “What were you just thinking about?” Response options were: “1) Concentrating on doing the experiment” (Task-Related Thought); “2) Things related to the experiment” (Task-related Interference, involving analysis/evaluation of task performance); “3) Something in the environment” (External Distraction, involving external stimuli unrelated to the task); “4) Something completely unrelated to the experiment, and unrelated to things in the environment” (Task-Unrelated Thought, corresponding to mind wandering/daydreaming); “5) I wasn't thinking about anything. My mind was ‘blank’” (Blank Mind, involving conscious awareness without content; Jackson & Balota, 2012; Watts, MacLeod, & Morris, 1988). Participants responded with the appropriate key. These five response options were provided because it is important to distinguish mind wandering from other forms of mentation which may

show different associations with questionnaire and behavioural measures (Stawarczyk, Majerus, Maj, Van der Linden, & D'Argembeau, 2011; Stawarczyk et al., 2012). In addition, in the current study, if response option ‘4’ (Task-Unrelated Thought/TUT) was chosen, further questions were presented to enquire about whether the TUT occurred with or without awareness; whether the TUT was intentional or unintentional; and what the nature of the TUT content was (with options of: friends/family; a personal matter; body sensation; singing; a personal memory; a fantasy; a news story; other). Participants undertook the SART individually in a test room. In the practice session the task instructions and thought-probe categories were explained, and then the participant was left to complete the task. Data recording errors occurred for four participants, leaving N = 126 for all thought-probe and behavioural measures. Ethical approval for these studies was obtained from the Internal Review Board of the Institute of Psychology, Chinese Academy of Sciences. 2.4. Statistical analysis Reported descriptive statistics include the mean, standard deviation, and range for each scale. Reliability was assessed with Cronbach's alpha. In addition, the validity of measures of mind wandering, daydreaming frequency and mindfulness was indicated by correlations with the SART thought-probe responses and behavioural measures (cf. Stawarczyk et al., 2012). Pearson product-moment correlations were calculated, controlling for age and gender. Mediation analysis was conducted following the procedures recommended by Preacher and Hayes (2004, 2008). Reported p-values are for two-tailed tests, and without Bonferroni correction. While the probability of making Type I errors can be reduced with the application of Bonferroni corrections, they also increase the chance of Type II errors, and there is disagreement about when and how they should be used (Feise, 2002; Nakagawa, 2004). Small, medium and large effect sizes of Pearson's r may be indicated by values of 0.10, 0.30 and 0.50 respectively, and to obtain a power of 0.80 at the 5% significance level for a medium effect size, the minimum sample size is 85 (Cohen, 1992). 3. Results 3.1. MCQ-30 factor structure CFA (AMOS v.17; maximum likelihood estimation, with co-varied factors) was undertaken on the combined MCQ-30 data (Study 1 + Study 2; total N = 384). Adequate model fit may be indicated by relative/normed Chi-square (Chi-squared statistic divided by the degrees of freedom) b 5.0; CFI (Comparative Fit Index) N0.90; SRMR (standardised root mean square residual) b 0.08; and RMSEA (root mean square error of approximation) b0.08 (Brown, 2006; Hooper, Coughlan, & Mullen, 2008). The CFA showed: relative/normed Chisquare = 2.558; CFI = 0.834; SRMR = 0.0721; RMSEA = 0.064 (90% confidence interval = 0.059 to 0.069). With 3/4 indices within acceptable ranges, the CFA gave some support for the five-factor structure of the MCQ-30, as reported by Wells and Cartwright-Hatton (2004). Inter-correlations between MCQ-30 subscales (for the combined data, N = 384) ranged − 0.004 (MCQ1/MCQ3) to 0.513 (MCQ3/ MCQ5). Test-retest coefficients ranged 0.557 (MCQ2) to 0.715 (MCQ1). Wells and Cartwright-Hatton (2004) reported MCQ-30 subscale inter-correlations ranging 0.22 to 0.60, and retest coefficients ranging 0.59 to 0.87 (for a mean interval of 34.14 days). 3.2. Descriptive statistics Table 1 shows the mean, standard deviation, range, and Cronbach's alpha for each scale (Study 1 and Study 2 data combined). MCQ-30 subscales showed reasonable/good alpha values, ranging 0.722 to 0.845 (cf., 0.72–0.93 reported by Wells & Cartwright-Hatton, 2004). Values of

R. Carciofo et al. / Personality and Individual Differences 107 (2017) 78–87

81

Table 1 Descriptive statistics (Study 1 and Study 2 data combined). Total N

Mean (SD)

Range (possible)

Cronbach's alpha

MCQ1 cognitive confidence

384 384

MCQ3 cognitive self-consciousness

384

MCQ4 uncontrollability/danger

384

MCQ5 need to control thoughts

384

MCQ-30 total score

384

Daydreaming frequency scale

372

Mind wandering scale

373

The Mindful Attention Awareness Scale-Lapses Only

378

6–24 (6–24) 6–24 (6–24) 6–24 (6–24) 6–24 (6–24) 6–24 (6–24) 34–106 (30−120) 13–60 (12–60) 17–60 (12–60) 26–67 (12–72) 15–50 (10–50) 10–48 (10–50) 18–58 (12–60) 0–13 (0−21)

0.763

MCQ2 positive beliefs

12.04 (3.872) 12.96 (3.809) 15.97 (3.583) 11.98 (3.996) 13.40 (3.965) 66.34 (12.452) 35.34 (8.950) 36.85 (6.763) 45.59 (6.066) 31.07 (5.703) 20.35 (7.031) 36.80 (6.543) 5.31 (2.214)

PANAS positive affect

a

368

PANAS negative affecta

368

Problem-solving daydreams scale

243

Pittsburgh Sleep Quality Index (global score)

368

0.845 0.779 0.799 0.722 0.862b 0.890 0.865 0.717 0.849 0.893 0.848 0.564

MCQ = MetaCognitions Questionnaire (30 item). PANAS = the Positive and Negative Affect Schedule. a The PANAS was administered twice in Study 2 (see Method). b For all 30 items; for the 5 subscales alpha = 0.653.

alpha for the other scales were all N 0.700, except for the PSQI (0.564 for the seven components combined; cf. Cheng, Shih, Lee, et al., 2012). Of the 368 participants who provided complete PSQI data, 367 scored zero for component 6 (use of sleep medication). Consequently, this component was excluded from further analyses. 3.3. Correlational analysis Correlations for the main variables (Study 1 and Study 2 data combined), are shown in Table 2. MCQ1 (Cognitive Confidence), correlated r N 0.3 (ps ≤ 0.0005) with daydreaming frequency, mind wandering, and mindfulness (negatively); it also correlated r N 0.2 (ps ≤ 0.0005) with negative affect, PSQI component 7 (daytime dysfunction), and PSQI global score. MCQ4 (Uncontrollability/Danger), correlated r N 0.2 (ps ≤ 0.0005) with daydreaming frequency, mind wandering,

mindfulness (negatively), negative affect, PSQI component 2 (sleep latency), PSQI component 7 (daytime dysfunction), and PSQI global score. MCQ5 (Need to Control Thoughts) correlated 0.213 with daydreaming frequency, but 0.080 with mind wandering. MCQ2 (Positive Beliefs), and MCQ3 (Cognitive Self-consciousness) did not correlate with mind wandering or daydreaming frequency (all rs b 0.1). However, in Study 1, Problem-solving Daydreams correlated with MCQ2 (r = 0.218, p ≤ 0.01; N = 243), MCQ3 (r = 0.338, p ≤ 0.0005; N = 243), MCQ5 (Need to Control Thoughts; r = 0.199, p ≤ 0.01; N = 243), and MCQ-30 total score (r = 0.238, p ≤ 0.0005; N = 243). Replicating previous findings, Daydreaming Frequency (DF) and Mind Wandering (MW) were strongly correlated (Table 2; cf. Carciofo et al., 2014a; Singer & Antrobus, 1972), and negatively correlated with mindfulness (MLO; cf. Carciofo et al., 2014a; Mrazek et al., 2012;

Table 2 Correlations between scales (Study 1 and Study 2 data combined).

MCQ1 CC MCQ2 PB MCQ3 CS MCQ4 UD MCQ5 CT MCQ-30 total score DF MW MLO Positive Affect Negative Affect

DF

MW

MLO

Positive affecta

Negative affecta

PSQI C1

PSQI C2

PSQI C3

PSQI C4

PSQI C5

PSQI C7

PSQI global

0.330⁎⁎⁎ 0.028 0.095 0.409⁎⁎⁎ 0.213⁎⁎⁎ 0.336⁎⁎⁎

0.352⁎⁎⁎ −0.048 −0.015 0.296⁎⁎⁎

−0.332⁎⁎⁎ −0.035 −0.047 −0.345⁎⁎⁎ −0.202⁎⁎⁎ −0.301⁎⁎⁎ −0.417⁎⁎⁎ −0.433⁎⁎⁎

−0.147⁎⁎ 0.086 0.231⁎⁎⁎ −0.113⁎

0.223⁎⁎⁎ 0.081 0.090 0.394⁎⁎⁎ 0.233⁎⁎⁎ 0.320⁎⁎⁎ 0.409⁎⁎⁎ 0.309⁎⁎⁎ −0.308⁎⁎⁎

0.095 0.011 0.040 0.177⁎⁎ 0.068 0.123⁎ 0.080 0.098 −0.101 −0.106⁎ 0.236⁎⁎⁎

0.152⁎⁎ 0.022 −0.009 0.251⁎⁎⁎ 0.107⁎ 0.166⁎⁎ 0.231⁎⁎⁎ 0.194⁎⁎⁎ −0.183⁎⁎⁎ −0.139⁎⁎ 0.231⁎⁎⁎

0.124⁎ 0.140⁎⁎ 0.117⁎

0.069 −0.042 0.059 0.107⁎

0.166⁎⁎ 0.141⁎⁎ 0.086 0.146⁎⁎

0.052 0.076 0.168⁎⁎ 0.105⁎ −0.055 −0.081 0.082

0.090 0.195⁎⁎⁎ 0.094 0.104⁎ −0.137⁎⁎ −0.120⁎ 0.104⁎

0.202⁎⁎⁎ 0.058 0.071 0.249⁎⁎⁎ 0.173⁎⁎ 0.235⁎⁎⁎ 0.308⁎⁎⁎ 0.349⁎⁎⁎ −0.207⁎⁎⁎ −0.214⁎⁎⁎ 0.200⁎⁎⁎

0.234⁎⁎⁎ 0.088 0.093 0.306⁎⁎⁎ 0.191⁎⁎⁎ 0.285⁎⁎⁎ 0.284⁎⁎⁎ 0.272⁎⁎⁎ −0.197⁎⁎⁎ −0.204⁎⁎⁎ 0.305⁎⁎⁎

0.080 0.211⁎⁎⁎ 0.682⁎⁎⁎

0.064 0.031 −0.264⁎⁎⁎ −0.282⁎⁎⁎ 0.104⁎

−0.058

0.083 0.163⁎⁎ 0.192⁎⁎⁎ 0.059 0.027 0.042 −0.021 0.143⁎⁎

Correlations controlling for age and gender. Ns = 355–384. MCQ = MetaCognitions Questionnaire (30-item); CC = cognitive confidence; PB = positive beliefs; CS = cognitive self-consciousness; UD = uncontrollability and danger; CT = need to control thoughts; DF = daydreaming frequency; MW = mind wandering; MLO = The Mindful Attention Awareness ScaleLapses Only; PSQI = Pittsburgh Sleep Quality Index; C1–C7 = PSQI components 1–7. a Positive/Negative Affect measured with the Positive and Negative Affect Schedule (PANAS). ⁎ p ≤ 0.05. ⁎⁎ p ≤ 0.01. ⁎⁎⁎ p ≤ 0.0005.

82

R. Carciofo et al. / Personality and Individual Differences 107 (2017) 78–87

Stawarczyk et al., 2012). DF and MW correlated negatively with Positive Affect, and positively with Negative Affect (cf. Stawarczyk et al., 2012), while MLO showed the opposite pattern of correlations (cf. Brown & Ryan, 2003). Several components of sleep quality (PSQI) positively correlated with DF and MW (i.e., poorer quality sleep associated with more daydreaming/mind wandering), and with negative affect (cf. Carciofo et al., 2014b). Problem-solving Daydreams did not significantly correlate with PSQI components (all rs b 0.120; all ps N 0.05; N = 238), or with negative affect (r = 0.073, p N 0.2; N = 232; cf. Giambra & Traynor, 1978); however, it correlated with positive affect, r = 0.200 (p ≤ 0.01; N = 232). 3.4. Experimental thought-probes (Study 2) Study 2 included the SART task, allowing for testing of the associations between metacognitions and daydreaming/mind wandering using data from the online/in-the-moment thought-probe responses, as an alternative measure to the retrospective, trait MW and DF questionnaires. Associations with SART behavioural measures were also tested, including percentages of hits and errors of commission, and the reaction-time coefficient of variation, i.e., the standard deviation of all of a participant's reaction times to SART stimuli, divided by the mean of all their reaction times; higher variation may indicate attention lapses (Cheyne, Solman, Carriere, & Smilek, 2009). Correlations between thought-probe responses, SART behavioural measures and questionnaire measures of mind wandering, daydreaming frequency and mindfulness are shown in Table 3. Task-related thought (TRT) negatively correlated with task-related interference, external distraction, task-unrelated thought (TUT) and blank mind. TRT was positively correlated with SART hits, and negatively correlated with SART errors and the coefficient of variation, while TUT negatively correlated with hits, and positively correlated with errors (rs N 0.1). These findings support the validity of the thought-probe responses, with more reported task focus associated with better task performance (cf. Stawarczyk et al., 2011). Furthermore, TUT was positively correlated with the DF and MW scales, and negatively correlated with the MLO (mindfulness) scale, while TRT showed the opposite patterns (Table 3; cf. Mrazek et al., 2012; Stawarczyk et al., 2012). DF and MW did not correlate with external distraction, although both correlated with reports of blank mind (as did TUT; Table 3). The DF scale was associated with poorer SART performance, while the MLO scale was associated with better performance. TUT also positively correlated with negative affect (r = 0.202, p ≤ 0.05; N = 124), and negatively correlated with positive affect (r = −0.180, p ≤ 0.05; N = 124). Correlations between SART responses and MCQ and PSQI scales are shown in Table 4. Some consistency with the Mind Wandering and Daydreaming Frequency scales was shown in the significant positive

correlations (rs N 0.2) between task-unrelated thought (TUT; mind wandering/daydreaming) and MCQ4 (Uncontrollability/Danger), MCQ5 (Need to Control Thoughts), and MCQ-30 total score. TUT also correlated (r N 0.1) with MCQ1 (Cognitive Confidence). Blank mind (BM) showed significant positive correlations with MCQ1 (Cognitive Confidence) and MCQ-30 total score, while task-related thought (TRT) showed a significant negative correlation with MCQ1 (Cognitive Confidence). These correlations demonstrate a consistent pattern of higher MCQ-30 scores (more maladaptive metacognitive beliefs) being associated with less task focus (TUT or BM). Furthermore, correlations between MCQ-30 subscales and SART behavioural measures (Table 4) generally showed trends in the direction of higher MCQ-30 scores being associated with poorer SART performance: fewer hits, more errors, and a higher coefficient of variation. Task-unrelated thought also showed positive correlations (rs N 0.1) with all PSQI components, indicating an association with poorer sleep quality. Aspects of poorer sleep quality were also associated with impaired SART performance, and negatively correlated with TRT, but there were no significant correlations with the other thought-probe categories, except between PSQI component 1 (subjective sleep quality) and blank mind. For the 82 participants who reported at least one episode of TUT, and so answered the additional questions, significant MCQ-30 correlations were found for: unaware TUT with MCQ4, r = 0.298 (p ≤ 0.01), and with MCQ-30 total score, r = 0.259 (p ≤ 0.05); unintentional TUT with MCQ4, r = 0.246 (p ≤ 0.05); and intentional TUT with MCQ2, r = 0.303 (p ≤ 0.01). Correlations between MCQ-30 subscales and the responses for the content of TUT were all p N 0.05, except: MCQ2 with friends/family, r = 0.228 (p ≤ 0.05); and MCQ4 with fantasy, r = 0.389 (p ≤ 0.0005). 3.5. Mediation analyses Previous studies have shown that the association between mind wandering/daydreaming frequency (predictor) and negative affect/ psychological distress (criterion) may be mediated by mindfulness and encoding style (Stawarczyk et al., 2012), and by sleep quality (Carciofo et al., 2014b). The current research investigated whether aspects of metacognitive belief might also act as mediators in this association. This was done by including all potential mediators in the same model (Fig. 1), thereby providing “… one way to pit competing theories against one another within a single model” (Preacher & Hayes, 2008, p.881). As mind wandering, daydreaming frequency and task-unrelated thought generally correlated most strongly with MCQ1 (Cognitive Confidence) and MCQ4 (Uncontrollability/Danger) these were included as possible mediators, in addition to mindfulness (MLO), and sleep quality (PSQI global score). Following the approach recommended by Preacher and Hayes (2004, 2008), we can identify: the effect of the predictor (IV) on the

Table 3 Correlations between SART thought-probes, questionnaires and behavioural measures.

Mean (SD) Daydreaming frequency (DF) Mind wandering (MW) Mindfulness (MLO) % Task-related thought (TRT) % Task-related interference (TRI) % External distraction (ED) % Task-unrelated thought (TUT) % Blank mind (BM)

% TRT

% TRI

% ED

% TUT

% BM

% Hit

% Error

Coefficient of variation

50.40 (24.06) −0.379⁎⁎⁎ −0.378⁎⁎⁎ 0.203⁎

27.65 (17.45) 0.121 0.183⁎ −0.007 −0.620⁎⁎⁎

7.28 (11.71) 0.025 −0.093 −0.096 −0.353⁎⁎⁎ −0.073

8.11 (9.15) 0.382⁎⁎⁎ 0.394⁎⁎⁎ −0.256⁎⁎ −0.497⁎⁎⁎

6.57 (12.12) 0.271⁎⁎ 0.280⁎⁎ −0.111 −0.393⁎⁎⁎

0.047 −0.012

−0.172 −0.123 0.186⁎

99.34 (0.98) −0.075 −0.068 0.025 0.168 −0.059 −0.090 −0.182⁎ −0.030

30.27 (15.36) 0.192⁎ 0.068 −0.161 −0.291⁎⁎ 0.227⁎

0.23 (0.05) 0.112 0.010 −0.183⁎ −0.167 0.120 0.111 0.099 −0.020

Correlations controlling for age and gender. N = 126. ⁎ p ≤ 0.05. ⁎⁎ p ≤ 0.01. ⁎⁎⁎ p ≤ 0.0005.

0.049 0.105 0.128

R. Carciofo et al. / Personality and Individual Differences 107 (2017) 78–87

83

Table 4 Correlations between SART responses, metacognitive beliefs and sleep quality.

MCQ1 CC MCQ2 PB MCQ3 CS MCQ4 UD MCQ5 CT MCQ-30 total PSQI C1 PSQI C2 PSQI C3 PSQI C4 PSQI C5 PSQI C7 PSQI global

% Task-related thought

% Task-related interference

% External distraction

% Task-unrelated thought

% Blank mind

% Hit

% Error

Coef./Var.

−0.256⁎⁎ 0.005 0.044 −0.098 −0.058 −0.109 −0.021 −0.253⁎⁎ −0.236⁎⁎

−0.042 −0.092 −0.134 −0.164 −0.105 −0.157 0.036 0.099 0.160 0.014 0.052 0.105 0.128

0.064 −0.001 0.011 0.120 −0.018 0.053 0.063 0.080 −0.018 −0.014 0.117 0.089 0.093

0.198⁎ 0.099 0.066 0.279⁎⁎ 0.205⁎⁎ 0.251⁎⁎ 0.146 0.285⁎⁎

0.359⁎⁎⁎ 0.050 0.045 0.109 0.130 0.204⁎ −0.176⁎

0.087 0.101 −0.110 −0.137 −0.062 −0.035 −0.033 −0.024 0.043 −0.344⁎⁎⁎ −0.207⁎ −0.038 −0.131

0.176 0.034 0.129 0.151 0.191⁎ 0.199⁎ −0.076 0.081 0.231⁎⁎

0.012 −0.046 0.185⁎ 0.158 0.111 0.122 −0.051 0.002 −0.090 0.159 0.078 0.036 0.024

−0.104 −0.153 −0.230⁎⁎ −0.273⁎⁎

0.141 0.208⁎ 0.213⁎ 0.204⁎ 0.328⁎⁎⁎

0.072 0.150 0.043 −0.040 0.069 0.027

0.053 0.022 0.065 0.091

Correlations controlling for age and gender. N = 126. MCQ = MetaCognitions Questionnaire (30-item); CC = cognitive confidence; PB = positive beliefs; CS = cognitive self-consciousness; UD = uncontrollability and danger; CT = need to control thoughts. PSQI = Pittsburgh Sleep Quality Index; C1–C7 = PSQI components 1–7. Coef./Var. = reaction-time coefficient of variation. ⁎ p ≤ 0.05. ⁎⁎ p ≤ 0.01. ⁎⁎⁎ p ≤ 0.0005.

mediator/s (path a); the direct effect of mediator/s on the criterion/DV, while controlling for the IV (path b); the total effect of the IV on the DV, while excluding the mediator/s (path c; this need not be significant for mediation effects to occur); the direct effect of the IV on the DV, while controlling for the mediator/s (path c′); and the indirect effect of the IV through the mediator/s (path ab); see Fig. 1. A non-parametric bootstrapping procedure (Preacher & Hayes, 2004, 2008) tested the indirect mediation effects by taking 5000 re-samples from the data to establish 95% confidence intervals, in which the exclusion of zero indicates a significant effect. Age and gender (coded 0 = male; 1 = female) were included as control variables. The results for the final regression models are reported here, including beta values for paths b and c′ (beta values for paths a and c are available from the authors). Three models were tested, with either Daydreaming Frequency (DF scale), Mind Wandering (MW scale), or Task-Unrelated Thought (thought-probe TUT) as the predictor/IV. For each model, negative affect (PANAS negative affect scale) was the criterion/DV (Table 5). Firstly, in Model 1, with the combined data for the Daydreaming Frequency scale (Study 1 + Study 2; N = 351), endorsement of beliefs about the uncontrollability/danger of thoughts (MCQ4), poor sleep quality (PSQI global score), and reduced mindfulness (MLO), were all significant mediators (the 95% confidence intervals excluded zero). Daydreaming Frequency retained a significant direct effect. In Model 2, with the combined data for the Mind Wandering scale (Study 1 + Study 2; N = 351), MCQ4, PSQI global score, and MLO were again significant mediators. Mind Wandering retained a significant direct effect. In both Model 1 and Model 2, contrast analyses of the mediation effects showed no significant differences between MCQ4, PSQI global score and MLO (the 95% confidence intervals included zero). However, in Model 3 with Task-Unrelated Thought as the predictor/IV (Study 2, thought-

probe data; N = 124), only MCQ4 was a significant mediator (Table 5). Task-Unrelated Thought did not retain a significant direct effect. As a more stringent criterion, the models were tested again using 99% confidence intervals. With Daydreaming Frequency or Mind Wandering as the predictor, both MCQ4 and PSQI global score were significant mediators (with contrast analysis showing no significant difference between them), but mindfulness (MLO) was not. With Task-Unrelated Thought as the predictor no mediators reached significance. Alternative analyses (using 99% confidence intervals) were also undertaken to test whether other aspects of metacognitive belief or sleep quality might also have mediation effects. Firstly, MCQ2, MCQ3 and MCQ5 were each added, in turn, as a fifth mediator in the models tested above. None showed significant mediation effects. With Daydreaming Frequency or Mind Wandering as the predictor, only MCQ4 and PSQI global score were significant mediators (with contrast analysis showing no significant difference between them). In the models with Task-Unrelated Thought as the predictor, only endorsement of beliefs about the uncontrollability/danger of thoughts (MCQ4) was a significant mediator. Secondly, for sleep quality, separate analyses were conducted substituting, in turn, each PSQI component in place of PSQI global score. With Daydreaming Frequency or Mind Wandering as the predictor none of the PSQI components were significant; in each model only endorsement of beliefs about the uncontrollability/danger of thoughts (MCQ4) was a significant mediator. With Task-Unrelated Thought as the predictor (Study 2), none of the PSQI components were significant mediators. The only significant mediator in these alternative models was MCQ4, although in the model testing PSQI component 4 (sleep efficiency) none of the mediators showed significance. In summary, for the combined questionnaire data, endorsement of beliefs about the uncontrollability/danger of thoughts (MCQ4), poor sleep quality and reduced mindfulness were all significant mediators at the 95% confidence interval. Mindfulness did not retain significance at the 99% confidence interval. For the thought-probe data, only endorsement of beliefs about the uncontrollability/danger of thoughts (MCQ4) was a significant mediator at the 95% and 99% confidence intervals, although it did not reach significance in two of the models tested at the 99% confidence interval. 4. Discussion

Fig. 1. Multiple mediator model.

The current studies investigated how aspects of metacognition identified in the metacognitive approach to psychological disorder (Wells & Matthews, 1996; Wells, 2007), are associated with mind wandering/

84

R. Carciofo et al. / Personality and Individual Differences 107 (2017) 78–87

Table 5 Mediation analyses: daydreaming frequency, mind wandering and task-unrelated thought as predictor (IV), and negative affect as criterion (DV).

Model 1 Predictor/IV = daydreaming frequency β Mean estimatea 95% C.I.b Model 2 Predictor/IV = mind wandering

R (adjusted R2)

F (df, residual)

0.583 (0.326)

25.215⁎⁎⁎ (7, 343)

0.567 (0.308)

23.213⁎⁎⁎ (7, 343)

0.403 (0.112)

3.211⁎⁎ (7, 116)

β Mean estimatea 95% C.I.b Model 3 Predictor/IV = task-unrelated thought β Mean estimatea 95% C.I.b

IV

Mediator 1 (MCQ1)

Mediator 2 (MCQ4)

Mediator 3 (PSQI global)

Mediator 4 (MLO)

Age

Gender

0.209⁎⁎⁎ −0.004 0.179⁎⁎ 0.158⁎⁎ −0.0019 0.0564 0.0339 −0.0324/0.0313 0.0219/0.0982 0.0113/0.0671

−0.102⁎ 0.0343 0.0005/0.0734

−0.354⁎⁎⁎ −0.009

0.133⁎

−0.002 0.215⁎⁎⁎ 0.164⁎⁎ −0.0013 0.0648 0.0483 −0.0490/0.0441 0.0311/0.1145 0.0150/0.0977

−0.118⁎ 0.0542 0.0042/0.1124

−0.336⁎⁎⁎ −0.006

0.055

0.066 0.287⁎⁎ 0.078 −0.035 −0.089 0.0078 0.0470 0.0131 0.0054 −0.0137/0.0431 0.0095/0.1139 −0.0190/0.0615 −0.0255/0.0417

0.069

MCQ = MetaCognitions Questionnaire (30-item); MCQ1 = cognitive confidence; MCQ4 = uncontrollability and danger; PSQI global = Pittsburgh Sleep Quality Index, global score; MLO = The Mindful Attention Awareness Scale-Lapses Only. a (Unstandardised) mean indirect effect estimate (path ab). b Bias corrected 95% confidence intervals for the (unstandardised) indirect effect; significant effects (in bold) shown by the exclusion of zero. ⁎ p ≤ 0.05. ⁎⁎ p ≤ 0.01. ⁎⁎⁎ p ≤ 0.0005.

daydreaming/task-unrelated thought. The new translation of the 30item version of the MetaCognitions Questionnaire (MCQ-30; Wells & Cartwright-Hatton, 2004) showed sound psychometric properties, with CFA supporting the five-factor structure, and reasonable/good internal consistency and test-retest reliability. Overall MCQ-30 score and several subscales (in particular MCQ4, beliefs about the uncontrollability and danger of thoughts) were associated with negative affect, consistent with the reported MCQ/MCQ-30 associations with psychological disturbance, including anxiety, depression, and obsessions (Cartwright-Hatton & Wells, 1997; Wells, 2007; Wells & Cartwright-Hatton, 2004; Solem, Hagen et al., 2015). The questionnaire measures of mind wandering, daydreaming frequency, mindfulness, affect, sleep quality, and problem-solving daydreams also generally showed good internal consistency, and the expected inter-correlations, consistent with previous research. Associations with task-unrelated thought, assessed with the thought-probe, and associations with SART behavioural data, were also as expected and consistent with previous research. 4.1. Metacognition and mind wandering/daydreaming It was found that mind wandering and daydreaming frequency are associated with aspects of metacognition, as assessed with the MCQ30. Higher frequencies of mind wandering and daydreaming (assessed with retrospective questionnaires and online thought-probes) were positively correlated with having less cognitive confidence (MCQ1), and with more endorsement of belief in the uncontrollability and danger of thoughts (MCQ4); daydreaming frequency and task-unrelated thought also positively correlated with more endorsement of belief in the need to control thoughts (MCQ5). While causal inferences cannot be drawn from correlational data, these findings seem consistent with the theory that mind wandering episodes can result from executive control failures (McVay & Kane, 2010), the experience of which might influence metacognitive beliefs about the controllability of thoughts, and also influence cognitive confidence. The observed positive correlations (rs N 0.2) between MCQ4 (belief in the uncontrollability and danger of thoughts), and unaware and

unintentional task-unrelated thought would also seem consistent with this view. Thus, people who experience frequent mind wandering due to executive control failures may only become aware of the episodes sometime after they began, which might lead to the episodes being judged as having occurred unintentionally. Such experiences might reduce confidence in general cognitive functioning, and lead to beliefs that some thoughts might be uncontrollable, and/or that there is a need to control thoughts. Alternatively, it might be that having less confidence in general cognitive functioning, and having belief in the uncontrollability of thoughts, might induce less effort in attempting to control mind wandering/daydreaming. Consistent with the current findings, belief in the uncontrollability/ danger of thoughts (MCQ4) is associated with long periods of rumination (Cartwright-Hatton & Wells, 1997), and the frequency of daydreaming may be associated with a more external locus of control (Lester & Tarnacki, 1989). Furthermore, the correlations between daydreaming/mind wandering and MCQ4 (Uncontrollability/Danger), are noteworthy given that this aspect of metacognition is particularly associated with measures of anxiety and worry (Wells & CartwrightHatton, 2004), and with depression (Solem, Hagen et al., 2015). Mind wandering and daydreaming frequency were also correlated with negative affect, as reported in many other studies (e.g., Giambra & Traynor, 1978; Killingsworth & Gilbert, 2010; Smallwood et al., 2009). This correlation was found with more trait-consistent measures (questionnaires), and also a more state/in-the-moment measure (thought-probe). However, the Problem-Solving Daydreams scale was not significantly correlated with negative affect (cf. Carciofo et al., 2014b; Giambra & Traynor, 1978), but showed significant correlations with positive affect, MCQ2 (Positive Beliefs about Worry), MCQ3 (Cognitive Self-consciousness), and MCQ5 (Need to Control Thoughts). Furthermore, MCQ2 correlated with intentional task-unrelated thought, while (as noted above), MCQ4 (belief in the uncontrollability and danger of thoughts), correlated with unintentional task-unrelated thought. These findings emphasize that, while the overall frequency of mind wandering/daydreaming seems reliably associated with negative affect, it is important to recognise the different forms, and content, that mind wandering/daydreaming can have (Mason et al., 2013; McMillan et al.,

R. Carciofo et al. / Personality and Individual Differences 107 (2017) 78–87

2013; Singer, 1978). Much mind wandering/daydreaming seems to involve ongoing current concerns (Klinger, 1978), which, to the extent that these involve some degree of stress/anxiety, might contribute to the association with negative affect. However, the frequency of mind wandering can also be increased by inducing positive mood (Seibert & Ellis, 1991). Some forms of mind wandering or daydreaming are also associated with positive affect, and constructive processes such as creativity (Mason et al., 2013; McMillan et al., 2013; Mooneyham & Schooler, 2013), while daydreaming of family/friends is associated with more well-being and life satisfaction (Mar, Mason, & Litvack, 2012). Also, achievement-oriented daydreams are more associated with having an internal locus of control, while fear of failure, and past-oriented daydreams are more associated with having an external locus of control (Brannigan, Hauk, & Guay, 1991; Brannigan, Shahon, & Schaller, 1992). Further research, employing more extensive measures, is needed to more fully study how aspects of metacognition are related to different forms and content of mind wandering/daydreaming. 4.2. Associations with sleep quality and mindfulness Aspects of metacognition had positive correlations with PSQI components, showing a consistent pattern of more maladaptive metacognitive beliefs being associated with poorer sleep quality. These findings seem consistent with previous research indicating that metacognition is involved in sleep disturbance, such as insomnia (Harvey et al., 2005; Waine et al., 2009). The correlation between sleep latency (PSQI component 2) and beliefs about the uncontrollability/danger of thoughts (MCQ4) is consistent with evidence that insomnia is related to negative beliefs about intrusive thoughts prior to sleep (Waine et al., 2009). There is also evidence that sufferers of primary insomnia may have more positive beliefs about worry, i.e., the benefits of worrying before sleep (Harvey et al., 2005; Waine et al., 2009). In the current studies, MCQ2 (positive beliefs about worry) correlated with sleep duration (PSQI component 3; i.e., more positive beliefs were associated with less sleep), and with sleep disturbances (PSQI component 5), but did not correlate with sleep latency (PSQI component 2). However, the MCQ was not developed to address metacognitive beliefs about sleep, so further research on this topic would benefit from a more specific measure, such as the MetaCognitions Questionnaire-Insomnia (MCQ-I; Waine et al., 2009), and from a comparison of clinical and non-clinical participants. As reported by Carciofo et al. (2014b), several aspects of poor sleep quality were positively correlated with mind wandering/daydreaming frequency, as assessed with retrospective questionnaires. In the current research, positive correlations (all rs N 0.1, most N0.2) were also found between all assessed aspects of poor sleep quality and the thoughtprobe measure of task-unrelated thought (mind wandering/ daydreaming), while several negative correlations were found with task-related thought. In contrast, correlations with the other thoughtprobe categories (task-related interference, external distraction, and blank mind) were weaker/less consistent. These findings indicate that sleep quality has a specific association with mind wandering/ daydreaming frequency (sleep quality was not significantly correlated with the Problem-Solving Daydreams scale; cf. Carciofo et al., 2014b). Also, most of the assessed aspects of poor sleep quality were associated with negative affect. Finally, mindfulness was assessed with the 12-item Mindful Attention Awareness Scale-Lapses Only (MLO; Carriere et al., 2008), which is a shorter version of the 15-item Mindful Attention Awareness Scale (MAAS; Brown & Ryan, 2003). The MLO scale had significant negative correlations with MCQ1 (Cognitive Confidence), MCQ4 (Uncontrollability/Danger), and MCQ5 (Need to Control Thoughts), while weaker correlations were shown with MCQ2 (Positive Beliefs about Worry) and MCQ3 (Cognitive Self-consciousness). The same pattern was found by Solem, Hagen et al. (2015) for the MAAS. Mindfulness also negatively correlated with mind wandering, daydreaming frequency and task-

85

unrelated thought (cf. Carciofo et al., 2014a; Mrazek et al., 2012; Stawarczyk et al., 2012). 4.3. Mediation analysis Previous research has shown that mindfulness (Stawarczyk et al., 2012) and sleep quality (Carciofo et al., 2014b) may mediate the association between mind wandering/daydreaming frequency (predictor) and negative affect (criterion). In the current research cognitive confidence (MCQ1) and beliefs about the uncontrollability and danger of thoughts (MCQ4) correlated with mind wandering, daydreaming frequency and task-unrelated thought, and also with negative affect. The mediation analysis thus simultaneously tested these metacognitive dimensions, mindfulness and sleep quality (PSQI global score), allowing for comparison of their relative effects. It was found that endorsement of belief in the uncontrollability and danger of thoughts (MCQ4) was the most consistently significant mediator. This was found with the mind wandering and daydreaming frequency questionnaires as predictors, and with thought-probe-assessed task-unrelated thought as a predictor. These results suggest that the frequency of mind wandering/daydreaming may influence the metacognitive beliefs that are held (specifically about the uncontrollability and danger of thoughts), and that these beliefs might lead to an increase in negative affect. However, this is not the only mechanism by which mind wandering/ daydreaming frequency may be related to negative affect. Both daydreaming frequency and mind wandering retained significant direct effects. There was also some indication with the questionnaire measures (based on a larger sample size than was involved for the thought-probe analysis) that mind wandering/daydreaming may influence negative affect through poor sleep quality (cf. Carciofo et al., 2014b), and through reduced mindfulness (cf. Stawarczyk et al., 2012). Other mediators, such as encoding style, may also be important (Stawarczyk et al., 2012). Nevertheless, the consistent mediation effects shown by endorsement of belief in the uncontrollability/danger of thoughts indicate that metacognition is an important aspect in the association between mind wandering and negative affect. Although Stawarczyk et al. (2012) argued that factors such as the strategies used to regulate thought and emotion might intervene between mindfulness and psychological distress, it seems important to emphasize that the strategies employed by an individual are based on their metacognitive beliefs (Wells & Matthews, 1996; Wells, 2007). Furthermore, recent research indicates that metacognition might be more important than mindfulness for understanding psychological distress. Solem, Hagen et al. (2015) found that cognitive confidence (MCQ1), positive beliefs about worry (MCQ2), and belief in the uncontrollability and danger of thoughts (MCQ4) were significant predictors of depression severity (as was anxiety), while mindfulness (as assessed with the MAAS scale), was not. In addition, Solem, Thunes, Hjemdal, Hagen, and Wells's (2015) joint factor analysis of the MCQ-30 and Five Facet Mindfulness Questionnaire (FFMQ; Baer, Smith, Hopkins, Krietemeyer, & Toney, 2006) produced distinct metacognition and mindfulness factors. The Metacognition factor included MCQ-30 subscales plus negative loadings for the FFMQ facets of Non-judging of Inner Experience and Acting with Awareness. The Mindfulness factor consisted of the Observing, Describing and Non-reacting to Inner Experiences subscales of the FFMQ. Solem, Thunes et al. (2015) found that the Metacognition factor was a strong predictor of symptoms of depression, anxiety and obsessivecompulsive symptoms, while the Mindfulness factor was only a weak predictor for depression. They concluded “… that the extent to which mindfulness is linked to maladaptation largely reflects metacognition”, and that while being a distinctive/unique feature of mindfulness, “… focusing on present moment experience may well be the least important from the perspective of psychological wellbeing” (Solem, Thunes et al., 2015, p.8–9). Similarly, while acknowledging that mindfulness training

86

R. Carciofo et al. / Personality and Individual Differences 107 (2017) 78–87

may help in the treatment of anxiety disorders by promoting metacognitive processing (with more flexible thinking, less constrained by habitual thought processes), Wells (2002) argued that increasing mindfulness does not necessarily change specific negative patterns of belief and thought, which would still need to be addressed. While mind wandering/daydreaming are not typically associated with clinical significance (Klinger et al., 2009; Singer, 1966), the metacognitive approach to psychological disorder (Wells & Matthews, 1996; Wells, 2007), might also be insightful in understanding cases of impaired daily functioning due to maladaptive/pathological daydreaming (see Schupak & Rosenthal, 2009; Somer, 2002). Interventions in such cases might include a focus on changing aspects of metacognitive belief. 4.4. Limitations and future research Further research, including larger samples and more demographic diversity, could investigate whether metacognition is a mediator in the relationships between mind wandering and other variables, such as life satisfaction and self-esteem (Luo, Zhu, Ju, & You, 2016). Further investigation of associations with mindfulness is also required. Studies of the association between mind wandering/daydreaming and mindfulness (e.g., Carciofo et al., 2014a; Stawarczyk et al., 2012), like the current research, have measured mindfulness with the Mindful Attention Awareness Scale (MAAS; Brown & Ryan, 2003), or its reduced version (Carriere et al., 2008). However, Baer et al. (2006) found that MAAS items only loaded on the ‘acting with awareness’ facet of their fivefacet conceptualisation of mindfulness. The five FFMQ facets show differential correlations with other variables (Baer et al., 2006), so might also vary in their associations with aspects of mind wandering or daydreaming. This could be investigated using the FFMQ and scales from the Imaginal Processes Inventory (Singer & Antrobus, 1972). As the nature and content of mind wandering/daydreaming can also moderate the relationship with negative affect (Mar et al., 2012; Mason et al., 2013; McMillan et al., 2013; Singer, 1978), future research could also comprehensively study how they may also interact with metacognition. Although the use of several questionnaire measures may be a methodological concern with the current research, the questionnaires showed consistent inter-correlations with the thought-probe measure of mind wandering and SART behavioural indices, demonstrating triangulation of results, supporting the validity of the measures (Schooler et al., 2011; Smallwood & Schooler, 2006; Stawarczyk et al., 2012). Nevertheless, the correlational nature of the data means that conclusions about causality cannot be drawn. Future studies could include interventions, to more fully test the importance of metacognitive beliefs in the association between mind wandering/daydreaming frequency and negative affect. 4.5. Conclusion Episodes of mind wandering or daydreaming, in their various forms, are frequently reported in the general population (McVay et al., 2009; Song & Wang, 2012), and negative affect is often associated with these experiences (Killingsworth & Gilbert, 2010). The current studies found that higher frequencies of mind wandering and daydreaming are positively correlated with dimensions of metacognition: having less cognitive confidence, more endorsement of belief in the uncontrollability and danger of thoughts, and also more endorsement of belief in the need to control thoughts. Furthermore, mediation analysis showed that endorsement of belief in the uncontrollability and danger of thoughts was a consistent mediator between mind wandering/ daydreaming frequency (predictor) and negative affect (criterion). Thus, the results of the current research show that the metacognitive approach to psychological disorder (Wells & Matthews, 1996; Wells, 2007) is useful for understanding the relationship between mind wandering/daydreaming frequency and negative affect. Furthermore, as

metacognitive frameworks have also been developed for mindfulness (Jankowski & Holas, 2014) and insomnia (Ong, Ulmer, & Manber, 2012), the broader study of mind wandering and daydreaming, in their various forms, might likewise benefit by being considered from the perspective of metacognitive theory. Acknowledgements This work was supported by the National Natural Science Foundation of China (grants 31470982 and 31200766), the Chinese Academy of Sciences (grant KJZD-EW-L04-4), and the Scientific Foundation of the Institute of Psychology, Chinese Academy of Sciences (grant Y4CX033008). These sources had no role in study design, data collection, analysis and interpretation, manuscript preparation, or decision to publish. References Baer, R. A., Smith, G. T., Hopkins, J., Krietemeyer, J., & Toney, L. (2006). Using self-report assessment methods to explore facets of mindfulness. Assessment, 13(1), 27–45. Brannigan, G. G., Hauk, P. A., & Guay, J. A. (1991). Locus of control and daydreaming. The Journal of Genetic Psychology, 152(1), 29–33. Brannigan, G. G., Shahon, A. J., & Schaller, J. A. (1992). Locus of control and time orientation in daydreaming: Implications for therapy. The Journal of Genetic Psychology, 153(3), 359–361. Brown, T. A. (2006). Confirmatory factor analysis for applied research. New York: Guilford Press. Brown, K. W., & Ryan, R. M. (2003). The benefits of being present: Mindfulness and its role in psychological well-being. Journal of Personality and Social Psychology, 84(4), 822–848. Buysse, D. J., Reynolds, C. F., III, Monk, T. H., Berman, S. R., & Kupfer, D. J. (1989). The Pittsburgh sleep quality index: A new instrument for psychiatric practice and research. Psychiatry Research, 28(2), 193–213. Carciofo, R., Du, F., Song, N., & Zhang, K. (2014a). Chronotype and time-of-day correlates of mind wandering and related phenomena. Biological Rhythm Research, 45(1), 37–49. Carciofo, R., Du, F., Song, N., & Zhang, K. (2014b). Mind wandering, sleep quality, affect and chronotype: An exploratory study. PLoS ONE, 9(3), e91285. http://dx.doi.org/ 10.1371/journal.pone.0091285. Carriere, J. S. A., Cheyne, J. A., & Smilek, D. (2008). Everyday attention lapses and memory failures: The affective consequences of mindlessness. Consciousness and Cognition, 17, 835–847. Cartwright-Hatton, S., & Wells, A. (1997). Beliefs about worry and intrusions: The metacognitions questionnaire and its correlates. Journal of Anxiety Disorders, 11(3), 279–296. Cheng, S. H., Shih, C. C., Lee, I. H., et al. (2012). A study on the sleep quality of incoming university students. Psychiatry Research, 197(3), 270–274. Cheyne, J. A., Solman, G. J. F., Carriere, J. S. A., & Smilek, D. (2009). Anatomy of an error: A bi-directional state model of task engagement/disengagement and attention-related errors. Cognition, 111(1), 98–113. Cohen, J. (1992). A power primer. Psychological Bulletin, 112(1), 155–159. Crawford, J. R., & Henry, J. D. (2004). The positive and negative affect schedule (PANAS): Construct validity, measurement properties and normative data in a large non-clinical sample. British Journal of Clinical Psychology, 43(3), 245–265. Feise, R. J. (2002). Do multiple outcome measures require p-value adjustment? BMC Medical Research Methodology, 2, 8. Flavell, J. H. (1979). Metacognition and cognitive monitoring: A new area of cognitive–developmental inquiry. American Psychologist, 34(10), 906–911. Giambra, L. M., & Traynor, T. D. (1978). Depression and daydreaming: An analysis based on self-ratings. Journal of Clinical Psychology, 34(1), 14–25. Giluk, T. L. (2009). Mindfulness, big five personality, and affect: A meta-analysis. Personality and Individual Differences, 47, 805–811. Grossman, P., & Van Dam, N. T. (2011). Mindfulness, by any other name…: trials and tribulations of sati in western psychology and science. Contemporary Buddhism, 12(1), 219–239. Harvey, A. G., Tang, N. K. Y., & Browning, L. (2005). Cognitive approaches to insomnia. Clinical Psychology Review, 25(5), 593–611. Hooper, D., Coughlan, J., & Mullen, M. (2008). Structural equation modelling: Guidelines for determining model fit. Electronic Journal of Business Research Methods, 6(1), 53–60. Howell, A. J., Digdon, N. L., Buro, K., & Sheptycki, A. R. (2008). Relations among mindfulness, well-being and sleep. Personality and Individual Differences, 45(8), 773–777. Huang, L., Yang, T., & Ji, Z. (2003). Applicability of the positive and negative affect scale in Chinese. Chinese Mental Health Journal, 17(1), 54–56 (Chinese language). Jackson, J. D., & Balota, D. A. (2012). Mind-wandering in younger and older adults: Converging evidence from the sustained attention to response task and reading for comprehension. Psychology and Aging, 27(1), 106–119. Jankowski, T., & Holas, P. (2014). Metacognitive model of mindfulness. Consciousness and Cognition, 28, 64–80. Killingsworth, M. A., & Gilbert, D. T. (2010). A wandering mind is an unhappy mind. Science, 330, 932.

R. Carciofo et al. / Personality and Individual Differences 107 (2017) 78–87 Klinger, E. (1978). Modes of normal conscious flow. In K. S. Pope, & J. L. Singer (Eds.), The stream of consciousness. New York: Plenum Press. Klinger, E., Henning, V. R., & Janssen, J. M. (2009). Fantasy-proneness dimensionalized: Dissociative component is related to psychopathology, daydreaming as such is not. Journal of Research in Personality, 43, 506–510. Lester, D., & Tarnacki, P. A. (1989). Frequencies of dreams and day-dreams and locus of control. Perceptual and Motor Skills, 69(3), 954. Liu, X., Tang, M., Hu, L., et al. (1996). Reliability and validity of the Pittsburgh sleep quality index. Chinese Journal of Psychiatry, 29(2), 103–107 (Chinese language). Luo, Y., Zhu, R., Ju, E., & You, X. (2016). Validation of the Chinese version of the mind-wandering questionnaire (MWQ) and the mediating role of self-esteem in the relationship between mind-wandering and life satisfaction for adolescents. Personality and Individual Differences, 92, 118–122. Mar, R. A., Mason, M. F., & Litvack, A. (2012). How daydreaming relates to life satisfaction, loneliness, and social support: The importance of gender and daydream content. Consciousness and Cognition, 21(1), 401–407. Marchetti, I., Koster, E. H., & De Raedt, R. (2012). Mindwandering heightens the accessibility of negative relative to positive thought. Consciousness and Cognition, 21(3), 1517–1525. Mason, M. F., Brown, K., Mar, R. A., & Smallwood, J. (2013). Driver of discontent or escape vehicle: The affective consequences of mindwandering. Frontiers in Psychology, 4, 477. http://dx.doi.org/10.3389/fpsyg. 2013.00477. McMillan, R., Kaufman, S. B., & Singer, J. L. (2013). Ode to positive constructive daydreaming. Frontiers in Psychology, 4, 626. http://dx.doi.org/10.3389/fpsyg.2013. 00626. McVay, J. C., & Kane, M. J. (2010). Does mind wandering reflect executive function or executive failure? Comment on Smallwood and Schooler (2006) and Watkins (2008). Psychological Bulletin, 136, 188–197. McVay, J. C., Kane, M. J., & Kwapil, T. R. (2009). Tracking the train of thought from the laboratory into everyday life: An experience-sampling study of mind wandering across controlled and ecological contexts. Psychonomic Bulletin & Review, 16(5), 857–863. Mooneyham, B. W., & Schooler, J. W. (2013). The costs and benefits of mind-wandering: A review. Canadian Journal of Experimental Psychology/Revue canadienne de psychologie expérimentale, 67(1), 11–18. Mrazek, M. D., Smallwood, J., & Schooler, J. W. (2012). Mindfulness and mind-wandering: Finding convergence through opposing constructs. Emotion, 12(3), 442–448. Nakagawa, S. (2004). A farewell to Bonferroni: The problems of low statistical power and publication bias. Behavioral Ecology, 15(6), 1044–1045. Ong, J. C., Ulmer, C. S., & Manber, R. (2012). Improving sleep with mindfulness and acceptance: A metacognitive model of insomnia. Behaviour Research and Therapy, 50(11), 651–660. Ottaviani, C., & Couyoumdjian, A. (2013). Pros and cons of a wandering mind: A prospective study. Frontiers in Psychology, 4, 524. http://dx.doi.org/10.3389/fpsyg.2013. 00524. Ottaviani, C., Shapiro, D., & Couyoumdjian, A. (2013). Flexibility as the key for somatic health: From mind wandering to perseverance cognition. Biological Psychology, 94, 38–43. Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behavior Research Methods, Instruments, & Computers, 36(4), 717–731. Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40(3), 879–891. 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(6), 747–758. Schooler, J. W., Smallwood, J., Christoff, K., Handy, T. C., Reichle, E. D., & Sayette, M. A. (2011). Meta-awareness, perceptual decoupling and the wandering mind. Trends in Cognitive Sciences, 15(7), 319–326.

87

Schupak, C., & Rosenthal, J. (2009). Excessive daydreaming: A case history and discussion of mind wandering and high fantasy proneness. Consciousness and Cognition, 18(1), 290–292. Seibert, P. S., & Ellis, H. C. (1991). Irrelevant thoughts, emotional mood states, and cognitive task performance. Memory & Cognition, 19(5), 507–513. Singer, J. L. (1966). Daydreaming. New York, NY: Random House. Singer, J. L. (1978). Experimental studies of daydreaming and the stream of thought. In K. S. Pope, & J. L. Singer (Eds.), The stream of consciousness. New York: Plenum Press. Singer, J. L., & Antrobus, J. S. (1972). Daydreaming, imaginal processes, and personality: A normative study. In P. W. Sheehan (Ed.), The nature and function of imagery (pp. 175–202). New York: Academic Press. Smallwood, J., & Andrews-Hanna, J. (2013). Not all minds that wander are lost: The importance of a balanced perspective on the mind-wandering state. Frontiers in Psychology, 4, 441. http://dx.doi.org/10.3389/fpsyg. 2013.00441. Smallwood, J., & Schooler, J. W. (2006). The restless mind. Psychological Bulletin, 132(6), 946–948. Smallwood, J., Fitzgerald, A., Miles, L. K., & Phillips, L. H. (2009). Shifting moods, wandering minds: Negative moods lead the mind to wander. Emotion, 9(2), 271–276. Solem, S., Thunes, S. S., Hjemdal, O., Hagen, R., & Wells, A. (2015). A metacognitive perspective on mindfulness: An empirical investigation. BMC Psychology, 3(1), 24. Solem, S., Hagen, R., Wang, C. E. A., Hjemdal, O., Waterloo, K., Eisemann, M., & Halvorsen, M. (2015). Metacognitions and mindful attention awareness in depression: A comparison of currently depressed, previously depressed and never depressed individuals. Clinical Psychology & Psychotherapy. http://dx.doi.org/10.1002/cpp.1983. Somer, E. (2002). Maladaptive daydreaming: A qualitative inquiry. Journal of Contemporary Psychotherapy, 32(2–3), 197–212. Song, X., & Wang, X. (2012). Mind wandering in Chinese daily lives – An experience sampling study. PLoS ONE, 7(9), e44423. http://dx.doi.org/10.1371/journal.pone.0044423. Stawarczyk, D., Majerus, S., Maj, M., Van der Linden, M., & D'Argembeau, A. (2011). Mindwandering: Phenomenology and function as assessed with a novel experience sampling method. Acta Psychologica, 136, 370–381. Stawarczyk, D., Majerus, S., Van der Linden, M., & D'Argembeau, A. (2012). Using the daydreaming frequency scale to investigate the relationships between mind-wandering, psychological well-being, and present-moment awareness. Frontiers in Psychology, 3, 363. http://dx.doi.org/10.3389/fpsyg.2012.00363. Stawarczyk, D., Majerus, S., & D'Argembeau, A. (2013). Concern-induced negative affect is associated with the occurrence and content of mind-wandering. Consciousness and Cognition, 22(2), 442–448. Waine, J., Broomfield, N. M., Banham, S., & Espie, C. A. (2009). Metacognitive beliefs in primary insomnia: Developing and validating the metacognitions questionnaire–insomnia (MCQ-I). Journal of Behavior Therapy and Experimental Psychiatry, 40(1), 15–23. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54(6), 1063–1070. Watts, F. N., MacLeod, A. K., & Morris, L. (1988). Associations between phenomenal and objective aspects of concentration problems in depressed patients. British Journal of Psychology, 79, 241–250. Wells, A. (2002). GAD, metacognition, and mindfulness: An information processing analysis. Clinical Psychology: Science and Practice, 9(1), 95–100. Wells, A. (2007). Cognition about cognition: Metacognitive therapy and change in generalized anxiety disorder and social phobia. Cognitive and Behavioral Practice, 14(1), 18–25. Wells, A., & Cartwright-Hatton, S. (2004). A short form of the metacognitions questionnaire: Properties of the MCQ-30. Behaviour Research and Therapy, 42(4), 385–396. Wells, A., & Matthews, G. (1996). Modelling cognition in emotional disorder: The S-REF model. Behaviour Research and Therapy, 34(11), 881–888.