Inattention and Hyperactivity Symptom Dimensions of ADHD Differentially Moderate the Relationship Between Concurrent Attention States and Affective Valence

Abstract

Background:

ADHD has been characterised by excessive mind wandering (MW), or thoughts unrelated to the task at hand, with recent findings indicating that ADHD is specifically associated with more unintentional, but not intentional, MW. These two types of MW are also differentially associated with affective well-being. Most existing studies in ADHD, however, mainly rely on retrospective reports of MW tendencies, which are susceptible to memory-related errors and biases. Further, most studies categorise participants based on overall levels of ADHD, instead of accounting for the spectrum and dimensional heterogeneity of ADHD, including inattention and hyperactivity symptom dimensions. Our study aimed to address the knowledge gap regarding the relationship between different types of MW and affective well-being, across different symptom dimensions of ADHD.

Methods:

We used ecological momentary assessment to capture participants’ momentary attention state (on-task, intentional MW, or unintentional MW) and affective valence, six times daily for 7 days. Using linear mixed-effects modelling to account for inter-individual variance, we tested whether inattention and hyperactivity symptom dimensions of ADHD differentially moderate the relationship between attention states and affective valence.

Results:

We found that higher levels of inattention symptoms predicted more negative affect during intentional MW compared to on-task attention; in contrast, higher levels of hyperactivity symptoms predicted more positive affect during intentional MW compared to on-task attention.

Discussion:

Together, our results indicate that intentional MW moderates opposing effects of inattention and hyperactivity ADHD symptoms on affective valence. Our findings suggest that intentional MW – and not just unintentional MW – may also play a role in affective or behavioural outcomes associated with ADHD symptomatology, and highlight the importance of considering the heterogeneity of ADHD symptomatology, as well as the distinction between intentional and unintentional MW, in future ADHD research.

Introduction

If you often find yourself thinking about a loved one or your last vacation while working, you are not alone. This ubiquitous human experience is conceptualised as a shift in attention away from the current task to one’s own internal thoughts, and is often referred to as mind wandering (MW; Smallwood & Schooler, 2015). Importantly, past research suggests that MW constitutes as much as one-third of our waking thoughts (Kane et al., 2007; Kawashima et al., 2023; Killingsworth & Gilbert, 2010; McVay et al., 2009), and at least certain forms of MW (such as when it occurs unintentionally) are robustly correlated with elevated negative affect (Seli et al., 2019). The majority of studies that have characterised the relationship between different types of MW and affective well-being thus far have been implemented in the general population (see Kam et al., 2024 for review). Recent research, however, has brought to light the importance of understanding this relationship in people with clinical conditions such as ADHD (Bozhilova et al., 2018; Kucyi et al., 2023). The current study therefore aims to examine the relationship between MW and affective well-being in ADHD.

Both theoretical and empirical work support a critical distinction between MW with and without intention (Seli et al., 2016). Specifically, intentional MW is engaged consciously and deliberately; for example, one may decide with intention to momentarily let their mind wander away from a mundane task, to brainstorm ideas for the next major project. In contrast, unintentional MW occurs spontaneously, often reflecting failures in attentional control; for example, while watching a news story about Berlin, one’s mind may unintentionally wander to thinking about friends who live in Germany. Empirical evidence supporting this distinction has found that unintentional MW is more strongly associated with symptoms of depression, anxiety, and negative affect compared to intentional MW, although intentional MW is also associated with negative affect (Kam et al., 2024; Seli et al., 2019). The distinction between intentional and unintentional MW has also been observed with respect to ADHD, such that greater ADHD symptoms have been associated with higher rates of unintentional, but not intentional MW (Arabacı & Parris, 2018; Lanier et al., 2021; Seli, Smallwood, et al., 2015). One possible explanation for the increase in unintentional MW is that it results from the impaired attentional control that characterises ADHD (Friedman-Hill et al., 2010). This is consistent with a review proposing that unintentional MW, but not intentional MW, may be a potential endophenotype underlying functional impairments associated with ADHD (Bozhilova et al., 2018).

However, a frequently cited study reporting a relationship between ADHD and unintentional MW (but not intentional MW), had characterised overall ADHD symptoms only (without discriminating between different symptom dimensions of ADHD), and their relationship with the frequency of intentional versus unintentional MW (Seli, Smallwood, et al., 2015). Consequently, it remains unclear to what extent unintentional MW is uniquely associated with any/all symptom dimensions of ADHD. To our knowledge, only one past study (Arabacı & Parris, 2018) has looked at the relationships between separate symptom dimensions of ADHD, and the frequency of intentional versus unintentional MW. Both studies provide evidence for a continuous relationship between ADHD symptom-levels and levels of intentional/unintentional MW within a general sample with non-diagnosed/non-clinical ADHD (Arabaci & Parris, 2018; Seli, Smallwood, et al., 2015). Finally, it is important to note that the frequency/occurrence of MW may not be the only relevant outcome from a neuropsychological perspective. MW often constitutes an extended period of maintenance of this attentional state (associated with self-generated thought), which is thought to involve various cognitive constructs and processes, including the affective content of thought during MW (Smallwood, 2013; Smallwood & Andrews-Hanna, 2013), and working memory (Levinson et al., 2012) – a construct of particular relevance to ADHD (Alderson et al., 2013; Martinussen et al., 2005). However, relatively little is known about the potentially differential effects of specific symptom dimensions of ADHD on cognitive constructs/processes during intentional/unintentional MW, other than their frequency. Investigating potentially differential effects of ADHD symptom dimensions, on one such construct – affective valence of thought during intentional/unintentional MW – was thus the primary focus of this study.

The Diagnostic and Statistical Manual of Mental Disorders (DSM)-5 characterises two main “clusters” of symptoms associated with ADHD, namely “inattention” and “hyperactivity-impulsivity,” whereby individuals may demonstrate symptoms from either or both symptom-clusters, to varying extents. Both theoretical and bio-medical research (Greven et al., 2011; Marcus & Barry, 2011) now provides sufficient evidence suggesting that the symptomatology of ADHD is multi-dimensional, whereby these clusters of partially independent ADHD symptoms can be characterised in the form of individual symptom dimensions. In general, the inattention dimension has been characterised by symptoms reflecting distractibility and forgetfulness, whereas the hyperactivity-impulsivity dimension has been characterised by symptoms reflecting physical and mental restlessness. These symptom dimensions have been found to be dissociable with respect to their persistence across age, associated functional impairments, and genetic heritability (Asherson, 2012; Garner et al., 2013; Greven et al., 2011; Marcus & Barry, 2011; Nikolas & Burt, 2010; Willcutt et al., 2012). Specifically, inattention symptoms tend to be persistent across the lifespan, whereas hyperactivity-impulsivity symptoms typically subside with age; yet, hyperactivity-impulsivity symptoms during childhood predict the appearance of inattention symptoms during adolescence, but not vice versa (Greven et al., 2011). Further, several lines of research suggest that measuring inattention and hyperactivity-impulsivity symptom dimensions separately is diagnostically and prognostically important, as they offer unique predictive information for functional outcomes with high discriminant validity (Kuntsi et al., 2014; Marcus & Barry, 2011). For example, inattention symptoms are more strongly associated with academic impairment and internalising psychopathology such as social withdrawal, whereas hyperactivity-impulsivity symptoms are more associated with externalising psychopathology such as adventure-seeking and social aggression (Garner et al., 2013). Finally, studies of genetic heritability also suggest different patterns of familial inheritance for inattention and hyperactivity-impulsivity symptom dimensions, with non-additive genetic effects being higher for inattention compared to hyperactivity-impulsivity, and additive genetic effects being higher for hyperactivity-impulsivity compared to inattention (Nikolas & Burt, 2010). However, past evidence also suggests considerable overlap between the genetic determinants and functional outcomes associated with both inattention and hyperactivity-impulsivity symptom dimensions of ADHD (Kuntsi et al., 2014), and researchers have highlighted the need for more studies contrasting the developmental, etiological, and psychological correlates of distinct symptom dimensions of ADHD. The aims of this study, investigating potentially dissociable relationships between separate symptom dimensions of ADHD, and cognition during intentional versus unintentional MW, are particularly relevant in this context.

Notably, although inattention and hyperactivity-impulsivity are the most well-characterised dimensions of ADHD, and most clearly aligned with the DSM-5 characterisation of ADHD symptomatology as two-dimensional, the exact dimensional structure of ADHD symptoms, especially in adult ADHD, remains debated. Indeed, all of two-factor (inattention, hyperactivity-impulsivity), three-factor (inattention, hyperactivity, impulsivity), and four-factor (inattention, hyperactivity, impulsivity, self-concept problems) symptom structures have found different degrees of support (Erhardt et al., 1999; Marcus & Barry, 2011; Martel et al., 2010; Park et al., 2018). In addition, a growing body of researchers have started to characterise a potential bi-factor latent structure of ADHD symptoms, with one general factor, and between two to three specific factors (Callahan & Plamondon, 2019; Martel et al., 2010). Based on current evidence, it seems likely that there may exist at least three partially independent symptom dimensions of ADHD: inattention, hyperactivity, and impulsivity. Of particular relevance to this study, the dimension of impulsivity is thought to be incompletely characterised by the DSM-5, as it includes only symptoms reflecting socio-motor disinhibition, like often blurting out answers or completing peoples’ sentences without waiting for them to finish, and often interrupting or intruding on others’ conversations or activities without asking for permission first. However, cumulative research suggests that impulsivity in ADHD most likely manifests in the form of symptoms reflecting both social and emotional dysregulation, such as unpredictable mood, irritability, and short temper (Callahan & Plamondon, 2019; Erhardt et al., 1999). For the purposes of this study, therefore, we decided to limit our investigation to the inattention and hyperactivity symptom dimensions of ADHD only, assessed using the Conners’ Adult ADHD Rating Scale (CAARS), as the overall dimensional structure of impulsivity symptoms as characterised by the CAARS and DSM-5 are not well-aligned. It is important to note, therefore, that the hyperactivity symptom dimension, and its effects as investigated in the present study, may not be sensitive to unique effects of the impulsivity symptom dimension, and should be cautiously considered if comparing these results with effects of the broader hyperactivity-impulsivity dimension, as has been dominantly characterised in past research.

Additionally, although traditionally considered a neurodevelopmental disorder that only impacts children, accumulating evidence suggests that ADHD is also highly prevalent in adults, who show a distinction between inattention and hyperactivity symptom dimensions (Callahan & Plamondon, 2019). Recent studies on adults with ADHD have reported that clinical levels of both inattention and hyperactivity are associated with more negative affective well-being (Faraone et al., 2015), as well as higher levels of MW (Biederman et al., 2019; Mowlem et al., 2019). Further, MW in general has itself been associated with lower levels of well-being in both neurotypical and ADHD populations (Kam et al., 2024; Mowlem et al., 2019). This makes it likely that small increments in ADHD symptoms may precipitate large drops in affective well-being through excessive MW. To date, however, little is known about the potentially differential relationships between ADHD symptomatology, in the adult population, and affective well-being during intentional versus unintentional MW.

Finally, past studies with ADHD populations have most often employed retrospective self-report questionnaires and diagnostic interviews, to assess the frequency of MW and (dys)regulation of affective states overall (Faraone et al., 2015; Mowlem et al., 2019; Seli, Smallwood, et al., 2015; Surman et al., 2015). Notably, the reliability of such retrospective reports depends crucially on accurate long-term memory recall, which is known to be compromised in ADHD (Sharma et al., 2021). Moreover, such retrospective methods often lead to a single data point for each individual and may not be sensitive to more temporally precise relationships between highly dynamic psychological constructs such as attention and affective states. One methodological approach that overcomes these limitations is ecological momentary assessment (EMA), which can reliably capture ongoing mental experiences in naturalistic settings (Moskowitz & Young, 2006; Shiffman et al., 2008). EMA involves prompting participants to report their experiences in-the-moment during their daily life, often multiple times a day for multiple days, weeks, or even months. This approach circumvents inaccuracies associated with memory errors and recall biases inherent to retrospective reports (Kuppens et al., 2022; Shiffman et al., 2008). Further, due to its repeated sampling (within-subjects) design, EMA is considered to be an ideal method for characterising both stable between-subject variance, and transient within-subject variance, in psychological constructs such as affective valence and attention which fluctuate throughout the day.

The current study therefore aimed to address whether inattention and hyperactivity ADHD symptom dimensions moderate the relationship between concurrent attention states and affective valence using linear mixed effects modelling applied to an EMA dataset. In a set of preliminary analyses, we first examined whether the inattention and hyperactivity symptom dimensions of ADHD are differentially associated with the relative frequencies of attention states (on-task, intentional MW, or unintentional MW). Based on existing literature that examined general levels of MW in ADHD (Arabacı & Parris, 2018; Bozhilova et al., 2018; Seli, Smallwood, et al., 2015), we hypothesised that higher levels of ADHD symptoms (inattention and/or hyperactivity) would be associated with more unintentional MW.

We then addressed our primary aim, concerning whether inattention and hyperactivity symptom dimensions differentially moderate the relationship between attention states and affective valence. Given past findings of an association between unintentional MW and negative well-being (Kucyi et al., 2023; Lanier et al., 2021), we hypothesised that higher levels of inattention and hyperactivity symptoms will predict more negative affect during unintentional MW specifically. In light of the knowledge gap in our understanding of the symptomatology and diagnostic criteria for ADHD in older age (Brod et al., 2012; Callahan & Plamondon, 2019), we recruited adults across a wide age range to increase the generalisability of our findings beyond young adulthood.

Methods

Participants

A total of 115 participants took part in the study, and our analyses included data from a final sample of N = 101 participants across a relatively large age-range: 19 to 79 years old (M = 39.68, SD = 14.31, Mdn = 38, IQR = 19), consisting of 80 females (age: M = 38.48, SD = 13.76) and 21 males (age: M = 44.29, SD = 15.74). We explain the rationale for data exclusion as part of Statistical Analyses below. Participants were recruited via email, from a pre-existing registry of individuals with formally diagnosed or suspected ADHD, in the local community in and around the city of Calgary (Alberta, Canada). The registry is part of an ongoing project titled Registry for Adult ADHD Research (RADAR) at the University of Calgary, led by author B.L.C. It serves as a pool of individuals who may be interested and eligible to participate in ADHD research studies. The gender-imbalance in the present study sample is line with the current composition of the registry, with 81% of enrollees being women, and 15% being men (4% “other” or “prefer not to say”). This is consistent with greater willingness of women to volunteer for health research relative to men (e.g., Glass et al., 2015). Enrollees’ ages ranged from 18 to 86 years, with a mean age of 39.5 years (closely resembling the present study’s sample distribution). Interested participants were asked to complete an online screening form, and individuals who fulfilled the following inclusion criteria based on their responses to this screening form were invited to take part in the study: individuals who (a) had access to an email account, cell phone, and a personal computer with a webcam and microphone, (b) were fluent in English, (c) had normal or corrected-to-normal eyesight, and (d) were between 18 and 80 years of age. Criterion (a) was necessary because this study was conducted online, and because we needed to send text messages with a survey link to their phone, as explained in the section Ecological Momentary Assessment below. No additional diagnostic testing for ADHD was undertaken, nor was formal diagnosis an inclusion or exclusion criterion. Participants were, however, asked for this information (see Demographic Survey and Clinical Questionnaires below). All participants provided informed consent, and were compensated with 40 Canadian Dollars in the form of an electronic gift card, for taking part in the study. The study was approved by the University of Calgary’s Conjoint Faculties Research Ethics Board (REB21-1513). The present study is part of a larger project that had other components aimed to address separate research questions. For transparency, we report all components in Supplemental Table S1.

Materials and Procedures

Participants who fulfilled all inclusion criteria met an experimenter online via Zoom (Zoom Communications Inc.). They first completed a 20-min experimental task designed to address a separate research question. Next, participants were provided detailed instructions about the daily EMA surveys. To ensure participants had a clear understanding of the survey items, the experimenter provided definitions and examples of each item, and supervised participants as they responded to a practice survey (which was not included in analyses). After the Zoom session, participants were emailed demographic and clinical questionnaires that they completed online, and the EMA protocol began 2 days after they completed these surveys. All questionnaires and surveys were implemented in Qualtrics (Qualtrics Inc.).

Demographic Survey and Clinical Questionnaires

The demographic survey asked participants to report their age (in years), their gender, as well as several other clinical and socio-economic variables, including whether they have ever been professionally diagnosed with ADHD, and whether they are currently taking medication for ADHD. Participants also completed the Conners Adult ADHD Rating Scales-Self-Report: Short-Version (CAARS-S:S), which assesses the severity of participants’ ADHD symptoms along several dimensions. As mentioned before, our main focus was on the inattention and hyperactivity symptom dimensions of ADHD only. Overall, the CAARS has been normatively developed to capture a four-factor latent structure of ADHD symptoms, which includes inattention, hyperactivity, impulsivity (emotional dysregulation), and problems with self-concept. The normative fit for this four-factor model was very good for a large clinical ADHD sample within the North American population, as well as showed very good fit for both sexes (Erhardt et al., 1999). Further, the scale also achieved acceptably good levels of specificity (87%) and sensitivity (82%) towards clinical diagnosis (Erhardt et al., 1999). The self-report short form version of the CAARS – used for the present study specifically – has been adapted from the larger four-factor CAARS model of ADHD, and has also been shown to have high internal reliability and measurement-invariance to sex within the general population in most dominantly English-speaking countries (Wu et al., 2023), making it a suitable tool for assessing dimensional ADHD symptom levels in reference to the general population within the “Western” cultural context.

Both subscales had good internal reliability in our sample (Cronbach’s α = .89 for inattention and .78 for hyperactivity). Raw scores were converted to T-scores for each subscale based on the participant’s gender and age, as specified in the user’s manual, and were used to operationalise the severity of respective ADHD symptom dimensions (Conners et al., 1999). Given that our main goal (see Primary Analysis below) was to assess ADHD effects on momentary affective valence (dependent variable) and that ADHD is thought to be significantly comorbidity with depression and anxiety (Michielsen et al., 2013), we included general indicators of affective well-being as covariates in the analyses to control for their overall effects at the participant-level. Specifically, we used the Patient Reported Outcomes Measurement Information System (PROMIS) depression (8a) and PROMIS anxiety (8a) scales. Both scales consist of eight items, and participants self-reported depression or anxiety indicators on a 5-point Likert scale; total scores across all items on each scale were operationalised as an index of depression and anxiety respectively. Both scales are considered to be clinically valid and reliable measures of negative affect in healthy and clinical populations (Cella et al., 2010; Schalet et al., 2016), and had excellent internal reliability in our sample (Cronbach’s alpha = .94 for both).

Ecological Momentary Assessment (EMA)

We used EMA to capture participants’ momentary attention state and affective valence in naturalistic settings. EMA is especially appropriate for a clinical population such as ADHD which has been associated with both short- and long-term memory impairments (Martinussen et al., 2005; Sharma et al., 2021; Skodzik et al., 2017), as it allows us to assess experiences in the moment, circumventing memory-related sources of error. Specifically in our study, participants were sent text messages, with links to an identical survey each time, at six pseudorandom timepoints every day for 1 week. All text messages were sent within a 10-hr time window based on participant’s self-reported waking hours. The survey was implemented in Qualtrics. Participants were asked to respond to the items on this survey based on their momentary experience immediately before they received the text message. Each survey contained five items to address different research aims, two of which were used in the present study. These items were “Were your thoughts on-task, mind wandering without intention, or mind wandering with intention at the time you received the survey?” (response options: on-task, mind wandering without intention, and mind wandering with intention), and “Please rate how you felt when you received the survey” (response options: Likert-type scale from (1) extremely negative to (7) extremely positive). Participants were instructed to respond to the surveys as soon as possible after they noticed it arrive, at most within 30 min; if they could not respond within that timeframe for any reason, participants were instructed to skip that survey. This is a common methodological choice to maximise the likelihood that the EMA reports are reliable and accurate (Mills et al., 2018; Thiemann et al., 2022).

Statistical Analyses

Data Exclusion

Of the 115 total participants who took part in the study, 5 did not provide complete responses for the clinical and demographic variables of interest necessary for analyses. The remaining 110 participants were considered for further analyses. Next, we excluded all EMA responses completed more than 30 min after receiving the text message prompt (381 responses, 19% of total responses), as well as responses that took more than 5 min to complete (140 responses, 7% of total responses). As the median response time to complete each survey was 37.5 s, surveys that took more than 5 min to complete likely reflected participants concurrently attending to other tasks, thereby decreasing the reliability of EMA reports. We also excluded data from two participants (37 responses) because the record of when surveys were distributed to them was missing from the Qualtrics output. Further, in order to assign CAARS-S:S T-scores of ADHD symptoms to our participants, which depended on their age and gender (either male or female), we excluded seven additional participants (155 responses) who did not identify as either male or female. This gave us a final sample of N_EMA = 2041 EMA responses across N_Participants = 101 participants.

Preliminary Analyses

We first examined whether the severity of inattention and hyperactivity ADHD symptom dimensions were differentially associated with the odds of unintentional and intentional MW relative to being on-task, and the odds of intentional MW relative to unintentional MW. To this end, we fit three separate logistic mixed-effects regression models with attention state as the categorical outcome variable, CAARS inattention and hyperactivity scores as the fixed-effect predictors of interest, along with age, gender, and PROMIS depression and anxiety scores as potential covariates; participant-level random-intercepts were also included. In the three models respectively, we tested the effects of predictors on the probability of unintentional MW relative to the probability of being on-task, the probability of intentional MW relative to the probability of being on-task, and the probability of unintentional MW relative to the probability of intentional MW. For omnibus tests of significance, each model was compared with a corresponding null model with only the participant-level random-intercepts and the grand-intercept. False discovery rate (FDR) correction (Benjamini & Hochberg, 1995) for three tests was applied to maintain an overall FDR of α = .05 across the three omnibus tests, as well as across the three parallel estimates for all fixed-effect parameters, including inattention and hyperactivity, across the three models (all of which were significant, see Results). All continuous predictors were centred across participants.

Primary Analyses

To investigate the primary research question – whether the severity of inattention and hyperactivity ADHD symptom dimensions differentially moderated the effect of attention states on affective valence – we implemented linear mixed-effects models, as they can efficiently partition and attribute the variance in momentary affect to between-subject (such as ADHD symptoms and average mood) and within-subject (such as attention state) predictors, in line with current frameworks of inter-individual variance in affect (Kuppens et al., 2022). To determine which predictors to include in the models, we implemented stepwise regression analyses, that is, we compared models with and without the predictor(s) of interest, using likelihood-ratio tests, to detect significant changes in model deviance at an α-threshold of .05. Based on past recommendations (Diggle et al., 2002; Zuur et al., 2009), we followed top-down model-selection, starting with a full model that included all fixed-effect predictors of interest, including interactions: attention state (categorical; on-task, intentional MW, or unintentional MW), both CAARS Inattention and Hyperactivity subscale T-scores (continuous), and their interactions (attention state × inattention) and (attention state × hyperactivity), as well as potential covariates including age (continuous), gender (categorical, male or female), and PROMIS depression and anxiety raw scores (continuous). All continuous predictors were centred across participants to reduce multicollinearity. In this model, we first fit the significant random-effects, to determine the optimal random-effects structure. We found that including both random participant-level intercepts, as well as random participant-level slopes of attention states, predicted affective valence significantly better than including just random-intercepts ( $\chi$ ² (5) = 42.12, p < .001); we therefore retained both types of random-effects in the full model to test our interaction hypotheses. Supplemental Tables S8 and S9 report all parameter estimates of the omnibus model.

To directly examine our primary question, we tested the significance of the two interactions of interest: (attention state × inattention) and (attention state × hyperactivity), in predicting affective valence (while including all above-mentioned covariates). To do this, we first fit the full model with the two interaction terms and all covariates; we then examined whether each interaction term, when separately excluded from the full model, led to a significant decrease in the explained variance in affective valence, by implementing likelihood ratio tests. We interpreted no significant decrease in explained variance as indicating that the interaction term did not contribute unique predictive information. The two a-priori interaction tests were corrected for multiple comparisons by controlling the FDR at α = .05. Residual plots and estimated variance inflation factors for the full model (with both interaction terms) confirmed that the analyses were not subject to major concerns relating to normality, homoscedasticity, or multicollinearity.

Finally, to follow up on the significant interaction terms, we removed all covariates from the full model which were not significant predictors of affective valence, and implemented two sets of follow-up analyses. First, we tested the significance of the simple slopes, characterising the effect of the corresponding ADHD symptom dimension (inattention, hyperactivity, or both, based on which interactions were significant) on affective valence, for each level of attention state (on-task, intentional MW, and unintentional MW) separately. Second, we tested for pairwise differences (between attention states) in the simple slopes characterised as part of the first set of follow-up analyses. Both sets of follow-up analyses were corrected for using an FDR of α = .05 across three tests, within each family of follow-ups to a significant two-way interaction.

Supplementary Analyses

To complement the main analyses, we also implemented a third set of follow-up analyses. Specifically, we quantified the Cohen’s d effect sizes of pairwise differences in affective valence, between different attention states, at the low (M − SD), average (M) and high (M + SD) levels of the CAARS subscales corresponding to significant two-way interactions (as above). Given this analysis conveys similar information as the two follow-up analyses described above, these results are reported in Supplemental Table S3. We report uncorrected p-values throughout, to preserve the raw p-values for direct interpretation; however all tests involving multiple comparisons underwent FDR-correction before interpretation as described before, and effects which did not survive FDR-correction are mentioned in-text (along with the FDR-corrected p-value threshold), wherever applicable. Finally, to characterise the clinical generalisability of our results, we also carried out supporting analyses after the primary analyses, whereby we included a “clinical diagnosis/medication” (hereafter “Clinical”) categorical covariate in all primary analyses if significant. Both 2- (Diagnosed/Not Diagnosed) and 3- (Diagnosed Unmedicated/Diagnosed Medicated/Not Diagnosed) level operationalisations of the Clinical covariate were explored, separately. All interactions of the Clinical covariate with all other predictors, including the two-way interactions of interest reported in this study, as well as its main effect on affect, were tested. We found that our primary two-way (mind wandering × ADHD symptom dimension) interactions of interest in the present study did not interact with the Clinical covariate. As such, results of these analyses are only reported in Supplemental Tables S10 to S13.

Results

Descriptive Statistics

Our final sample consisted of 2,041 EMA responses from 101 participants. The median CAARS T-scores for ADHD inattention and hyperactivity were 66 (M = 64.38, SD = 13.97, IQR = inter-quartile range = 23) and 60 (M = 59.96, SD = 10.22, IQR = 15) respectively, over the entire sample. Since T-scores are distributed with a mean of 50 and SD of 10 in the general population, this indicates participants in our sample represented a spectrum from mild to severe symptoms along both dimensions, with the group mean around one standard deviation above the general population mean (Conners et al., 1999). Participants in the final sample covered an age range of 19 to 79 years (Mdn = 38, IQR = 19), and included n = 80 females and n = 21 males.

Out of N = 101, participants, n = 58 had a formal diagnosis of ADHD (43 female, 15 male; CAARS Inattention T: Mdn = 71.00, IQR = 17.80; CAARS Hyperactivity T: Mdn = 66.00, IQR = 10.50), and n = 42 were undiagnosed (36 female, 6 male, CAARS Inattention: Mdn = 53.00, IQR = 13.80; CAARS Hyperactivity T-scores: Mdn = 52.00, IQR = 9.50); the diagnosis-status of one participant was unavailable. Of the 58 formally diagnosed participants, n = 40 indicated currently being on medication (31 female and 9 male; CAARS Inattention T-Scores: Mdn = 73.00, IQR = 18.75; CAARS Hyperactivity T: Mdn = 60.90, IQR = 10.25), and the other n = 18 indicated not being on medication (12 female and 6 male; CAARS Inattention T-Scores: Mdn = 70.50, IQR = 15.75; CAARS Hyperactivity T: Mdn = 62.10, IQR = 9.75). See Supplemental Table S2 for other descriptive statistics characterising the participants.

The mean affective valence across all 2,041 EMA responses was 4.38 (SD = 1.26), on a seven-point scale from (1) extremely negative to (7) extremely positive. When EMA responses were grouped within individuals, unintentional MW, intentional MW, and on-task attention were reported on 38.24% (SD = 19.80%), 21.64% (SD = 16.08%), and 40.13% (SD = 19.24%) of all valid EMA responses. The sample-wide mean affective valence for EMA responses, averaged within-individual, was 4.39 (range = 1.80–7.00, SD = 0.70). See Supplemental Table S4 for a complete correlation matrix of all continuous variables.

Preliminary Analyses

We first addressed whether ADHD symptom dimensions were differentially associated with the relative frequencies of occurrence of attention states. Only effects of inattention and hyperactivity symptoms are reported below for brevity; for all parameters of each model, see Supplemental Tables S5 and S6. For model-evaluation metrics, see Supplemental Table S7.

The model predicting the odds of intentional MW relative to on-task attention was significant ( $\chi$ ²(6) = 15.23, p = .019). We found that greater hyperactivity symptoms significantly predicted greater odds of intentional MW relative to on-task attention (OR = 1.03, 95% CI [1.01 1.05], p = .012), such that a unit increase in hyperactivity T-score was associated with a 3% increase in the odds of intentional MW relative to on-task attention. Inattention symptoms did not have a significant effect on the odds of intentional MW relative to on-task attention (p = .644). On average in our sample (grand-intercept), the odds of intentional MW were 57% less than the odds of being on-task (OR = 0.43, 95% CI [0.34 0.54], p < .001).

The model predicting the odds of unintentional MW relative to on-task attention was also significant ( $\chi$ ²(6) = 13.96, p = .031). However, none of the fixed-effect parameters (including inattention and hyperactivity symptom-effects) within the model were themselves significant predictors (ps > .094).

Finally, the model predicting the odds of unintentional MW relative to intentional MW was also significant ( $\chi$ ²(6) = 23.30, p < .001). Neither inattention nor hyperactivity symptoms had a significant effect on the relative frequencies of unintentional and intentional MW (ps > .392). However, on average in our sample (grand-intercept), the odds of unintentional MW were 144% higher than the odds of intentional MW (OR = 2.44, 95% CI [1.90 3.13], p < .001).

Primary Analyses

Main Effects on Affective Valence

The omnibus model indicated a significant negative effect of unintentional MW compared to being on-task (p = .016; but not intentional MW compared to being on-task, p = .454), as well as PROMIS anxiety scores (p = .020), on affective valence. Pairwise comparisons of affective valence across all attention state-pairs revealed that affective valence was only significantly lower during unintentional MW compared to on-task attention (p = .014), and no other attention state pairs were characterised by significantly different affective valence overall. See Table 1 for the pairwise differences in affective valence between all attention states.

Table 1.

Pairwise Differences in Affective Valence Between all Attention States.

Term	b	SE	95% CI	t-value (df)	p-value
On-Task – Unintent. MW	0.21	0.08	[0.04, 0.38]	−4.34 (2003)	.014
On-Task – Intent. MW	0.06	0.07	[−0.09, 0.21]	−0.78 (2004)	.395
Unintent. MW – Intent. MW	−0.14	0.08	[−0.31, 0.02]	2.81 (2007)	.069

Note. SE = standard error; CI = confidence interval. Unintent. MW = unintentional mind wandering, intent; MW = intentional mind wandering. Bolded p-values indicate significant effects after multiple comparisons correction at an FDR of α = .05.

Interactions Between ADHD Symptom Dimensions and Attention State in Predicting Affective Valence

We then examined the interactions between attention state and different ADHD symptom dimensions in predicting affective valence. We found that both the (attention state × inattention) interaction ( $\chi$ ²(2) = 8.36, p = .016) and the (attention state × hyperactivity) interaction ( $\chi$ ²(2) = 6.52, p = .039) were significant predictors of momentary affect. We then implemented two sets of analyses to follow-up on the significant interactions. Before performing the follow-up tests, the model was refit by dropping all covariates which were not significant predictors of affective valence. We first examined whether there was a significant association between the inattention or hyperactivity subscale scores and affective valence at each attention state separately through simple slopes; then we examined pairwise differences in these simple slopes between attention states.

For the CAARS inattention subscale, we found that symptom scores had a significant negative relationship with affective valence only during intentional MW (b = −0.02, SE = 0.01, 95% CI [−0.04 −0.00], p = .009), but not during unintentional MW or on-task attention. Further, the slope was significantly more positive for on-task attention compared to intentional MW (Δb = 0.02, SE = 0.00, 95% CI [0.00 0.03], p = .006); no other pairwise differences between slopes were significant.

For the CAARS hyperactivity subscale, we found that symptom scores had no significant association with affective valence during any of the attention states separately. However, in comparing pairs of attention states, we found that the slope of hyperactivity scores predicting affective valence was significantly more negative for on-task attention compared to intentional MW (Δb = −0.02, SE = 0.01, 95% CI [−0.04 −0.00], p = .014). Table 2 reports all parameter estimates of interest in the primary analysis. Figure 1 illustrates the interaction effects.

Table 2.

Simple Slopes of Inattention and Hyperactivity Symptom T-Scores Predicting Momentary Affective Valence for Each Attentional State Separately, and Pairwise Differences Between Them.

Term	b	SE	95% CI	t-value (df)	p-value
A. Simple slopes of inattention
Simple slopes
On-task	−0.00	0.01	−0.02, 0.21	−0.20 (102)	.840
Unintent. MW	−0.01	0.00	−0.02, 0.01	−1.15 (103)	.254
Intent. MW	−0.02	0.01	−0.04, −0.00	−2.71 (90)	.009
Pairwise differences in simple slopes
On-Task – Unintent. MW	0.00	0.01	−0.01, 0.02	0.82 (99.3)	.416
On-Task – Intent. MW	0.02	0.00	0.00, 0.03	2.83 (80.1)	.006
Unintent. MW – Intent. MW	0.01	0.01	−0.01, 0.03	1.72 (87.8)	.089
B. Simple slopes of hyperactivity
Simple slopes
On-task	−0.01	0.01	−0.03, 0.01	−0.97 (105.7)	.336
Unintent. MW	−0.00	0.01	−0.02, 0.02	−0.14 (104.1)	.891
Intent. MW	0.01	0.01	−0.01, 0.03	1.22 (91.2)	.227
Pairwise differences in simple slopes
On-Task – Unintent. MW	−0.01	0.01	−0.03, 0.02	−0.86 (93.0)	.395
On-Task – Intent. MW	−0.02	0.01	−0.04, 0.00	−2.52 (78.9)	.014
Unintent. MW – Intent. MW	−0.01	0.01	−0.03, 0.01	−1.41 (88.9)	.162

Note. These are parameter estimates for the follow-up analyses to the significant two-way (attention state × inattention) and (attention state × hyperactivity) interactions. b-values for simple slopes reflect the effect of Inattention (panel A) or Hyperactivity (panel B) T-scores on affective valence at the specific attentional state, while b-values for pairwise differences in simple slopes reflect the specific pairwise differences, between attentional states, in the aforementioned simple slopes. SE = standard error; CI = confidence interval; Unintent. MW = unintentional mind wandering, intent. MW = intentional mind wandering. Bolded p-values indicate significant effects after multiple comparisons correction at an FDR of α = .05.

Figure 1.

Simple slopes of inattention and hyperactivity symptom scores predicting affective valence (and pairwise differences in simple slopes between attention states).

Note. Shaded areas represent 95% confidence intervals of the effects of Inattention (panel A) and Hyperactivity (panel B) symptoms on affective valence (complete range: 1 to 7) at each attention state separately (simple slopes). b represents simple slopes, and Δb represents differences in these simple slopes (significant labelled effect: on-task – intentional MW, in both panels). *Represents significant effects at p < .05 after correction for multiple comparisons at an FDR of α = .05. M represents the across-participants mean of Inattention and Hyperactivity T-scores in panels (A) and (B) respectively.

Discussion

In the present study, we primarily investigated whether two major symptom dimensions of ADHD – inattention and hyperactivity – differentially moderated the relationship between attention states (including on-task attention, intentional, and unintentional MW) and affective valence, in everyday life. In a set of preliminary analyses, we first investigated whether ADHD symptom dimensions differentially predicted the relative frequencies of different attention states. We found that greater hyperactivity symptoms, but not inattention symptoms, were associated with an increased probability of intentional MW relative to on-task attention. Next, as part of the primary analyses, we found that greater inattention symptoms were associated with lower affect only during intentional MW, and that this effect was significantly more negative during intentional MW compared on-task attention states; in contrast, greater hyperactivity symptoms were associated with more positive affect during intentional MW compared to on-task attention. Taken together, our results suggest that the relationship between attention states and affective valence, as well as the relative frequencies of attention states, vary differentially as a function of individual symptom dimensions of ADHD. In line with past studies (Faraone et al., 2015; Garner et al., 2013; Nikolas & Burt, 2010; Seli, Carriere, & Smilek, 2015; Willcutt et al., 2012), our results highlight the importance of considering the spectrum and heterogeneity of ADHD symptomatology, as well as the distinction between intentional and unintentional MW, in future research.

Hyperactivity Symptoms Are Associated With More Frequent Intentional MW Relative to On-Task Attention

As part of the preliminary analyses, in examining ADHD symptom dimensions separately, we found that only hyperactivity symptoms had a significant effect on the relative frequencies of intentional MW and on-task attention, such that greater hyperactivity symptoms were associated with increased rates of intentional MW relative to on-task attention. There were no other significant effects of either inattention or hyperactivity symptoms on the relative frequencies of different attention states. This finding diverges from current frameworks characterising unintentional MW, but not intentional MW, as being uniquely associated with ADHD symptomatology, as well as negative well-being, in the ADHD population (Bozhilova et al., 2018; Seli, Smallwood, et al., 2015; Seli et al., 2016). In contrast to past studies that have mainly examined the frequency of MW reports using analyses of variance (ANOVAs), our analysis modelled the occurrence of categorical attention states using logistic regression, which is theoretically better suited to such frequency data. In addition to differences in statistical analyses, our study also differed from past studies (Arabacı & Parris, 2018; Seli, Smallwood, et al., 2015; Shaw & Giambra, 1993) in terms of the sampling approach. We used EMA, which circumvents memory-related errors and biases characterising retrospective reports, to capture in-the-moment experiences during daily life, whereas past studies have generally used questionnaire-based retrospective reports or probe-caught reports during laboratory-tasks (Arabacı & Parris, 2018; Seli, Smallwood, et al., 2015; Shaw & Giambra, 1993). Thus, our results may be more naturalistically valid, regarding the effects of ADHD symptoms on the relative probabilities of different types of MW during daily life.

Across our sample, we found that participants reported MW (either intentionally or unintentionally) a majority of the time (60%). Our results contrast with existing EMA studies in the general population, which suggest the overall proportion of MW is around 30% to 35% (Kane et al., 2007; Kawashima et al., 2023; McVay et al., 2009). The high levels of MW in our sample is in line with past studies which have reported positive associations between ADHD symptomatology generally (not differentiating between inattention and hyperactivity symptom dimensions) and the frequency of MW overall, both during laboratory tasks (Arabacı & Parris, 2018; Shaw & Giambra, 1993) as well as assessed using retrospective questionnaires (Mowlem et al., 2019; Seli, Smallwood, et al., 2015). Thus, our results provide support for excessive levels of MW in adults with relatively high levels of self-reported ADHD symptoms.

Importantly, our results may suggest a pattern of a general relationship between overall ADHD symptom levels (across dimensions) and MW, in addition to more specific relationships regarding individual symptom dimensions. A similar pattern was found by Arabacı and Parris (2018), employing easy and difficult sustained-attention to response tasks (SART), within-subject. They found hyperactivity symptoms were associated with more unintentional MW across conditions, while inattention symptoms were associated with more unintentional MW only during the difficult condition. In general, these results provide empirical backing for further research exploring bi-factor models of ADHD symptomatology. Finally, in the same vein as Arabacı and Parris (2018), we note that these results suggest unintentional MW may not be uniquely associated with all symptom dimensions of ADHD, and more research is needed for the purposes of replication and validation of these specific relationships.

Inattention and Hyperactivity Symptom Dimensions Differentially Moderate the Association Between MW and Affective Valence

As part of our primary analyses, we found that greater inattention symptoms were associated with lower affective valence only during intentional MW and that this effect was stronger for intentional MW than for on-task attention. Given existing frameworks have characterised unintentional MW to be primarily associated with negative well-being outcomes in ADHD (Bozhilova et al., 2018), our finding that greater inattention ADHD symptoms predicted greatest negative affect during intentional MW was therefore unexpected. One plausible explanation is that those with higher levels of inattention may engage in more maladaptive forms of intentional MW characterised by negative affect. This explanation is supported by a study which found a positive relationship between inattention symptoms and intrusive, ruminative thinking styles characteristic of depression, while controlling for hyperactivity symptoms (Jonkman et al., 2017). Consistent with this finding, studies have shown that both adults in the general population (Carriere et al., 2008) and adults with ADHD (Michielsen et al., 2018) tend to negatively evaluate themselves in response to attentional impairments (which are elevated in ADHD), often leading to low self-esteem and self-efficacy (Newark et al., 2016) as well as comorbid depression in both young and old adults (Michielsen et al., 2013). The above explanation is further supported by the observation that inattention ADHD symptoms are indeed more correlated than hyperactivity symptoms with internalising psychopathologies like depression, characterised by self-criticism and negative affect (Willcutt et al., 2012). Notably, we also found that inattention symptoms were more positively correlated with PROMIS depression scores than hyperactivity symptoms in our sample (see Supplemental Table S4). If negative self-evaluation is indeed the reason for the negative affect during intentional MW associated with ADHD inattention, then our results reinforce the importance of considering cognitive-behavioural interventions for ADHD (Safren et al., 2004) which focus on changing maladaptive patterns of self-appraisal that are known to arise out of functional impairments and lead to negative affect (Beaton et al., 2022).

In contrast to inattention symptoms, hyperactivity symptoms were not associated with affective valence at any attentional state individually; however, hyperactivity symptoms did predict a more positive association with affective valence during intentional MW compared to the on-task attention state. Specifically, hyperactivity symptoms did not predict affective valence during the on-task, intentional MW, and unintentional MW states separately, across participants. However, when comparing between attention states within-participant, we found that being on-task, relative to intentional MW, was associated with more negative affect, for individuals with higher levels of hyperactivity symptoms. One potential explanation could be related to the manifestation of both mental and physical restlessness in individuals with greater hyperactivity symptoms (Conners et al., 1999; Weyandt et al., 2003). It is possible that intentional MW may be associated with more positive affect compared to staying on-task for individuals with elevated hyperactivity symptoms, as MW may have been used as an internal strategy to cope with the restlessness. Indeed, individuals in the general population also demonstrate strategic adaptation of off-task thoughts in response to task demands (Rummel & Boywitt, 2014), and it is possible that the observed patterns here reflect the affective correlates of this process in our sample. Our finding that greater hyperactivity symptoms were associated with increased rates of intentional MW provides indirect support for this speculation.

Unintentional MW Is Specifically Associated With More Negative Affect

Finally, there was a negative main effect of unintentional MW on affective valence. Specifically, unintentional MW, but not intentional MW, was associated with significantly more negative affect compared to on-task attention. Our results replicate past reports of a unique relationship between unintentional MW and negative affect in the general population (Kam et al., 2024; Seli et al., 2019). One potential straightforward explanation of this association concerns a lack of awareness and/or control over one’s own thoughts. Specifically, it may be that participants reported negative affect directly because of becoming aware, due to the survey prompt, that their minds had wandered away from the task outside their own control and agency, such that the negative affect could be in response to negative self-evaluation following discovery of one’s failure of meta-awareness, which is linked to unintentional MW (Smallwood & Schooler, 2015). The self-perceived lack of control over one’s own thoughts or lack of agency over one’s actions have both been associated with decreased well-being in the general population (Kaiser et al., 2021; Kam et al., 2024) as well as ADHD (Bozhilova et al., 2018; Harpin et al., 2013). A related explanation concerns one’s evaluation of performance decrements associated with MW. In particular, ample research suggests that MW disrupts task performance across diverse contexts and task domains, including doing math (Randall et al., 2019), attending to lectures or learning in educational contexts (Szpunar et al., 2013; Wong et al., 2022), and driving (Gil-Jardiné et al., 2017). Therefore, recognising the disruption on task performance during unintentional MW may also lead to negative affect. Both potential explanations are in line with past studies suggesting that greater everyday task-errors caused by lapses in sustained attention positively correlated with both lower meta-awareness of one’s thoughts (which is conceptually associated with unintentional MW; but see Seli et al. (2017) and Smallwood (2013) for a more nuanced discussion) as well as more depressed mood and stress (Carriere et al., 2008; Cheyne et al., 2006). In summary, future research should investigate the causal mechanisms underlying the relationship between unintentional MW and negative affect, both in the clinical ADHD and general population.

Limitations and Future Directions

Several limitations of the present study need to be considered. First, the sampling of participants was cross-sectional; accordingly, our results should not be interpreted as evidence for incremental effects along a developmental trajectory of ADHD symptoms on attention state or affect. Indeed, our results highlight questions relating to within-participant relationships that can be addressed only with longitudinal study designs (Nigg, 2016). For example, future research can examine how ADHD symptom dimensions moderate the relationship between affective valence and attention states, longitudinally across time. The cross-sectional design also limits our understanding of the directionality of effects: we examined the relationship between momentary affect and attention state assessed concurrently, as such we cannot infer the directionality of this relationship. Accordingly, future studies can use an EMA design with probes paired closely in time (Poerio et al., 2013) to examine, for example, whether intentional or unintentional MW reported in an earlier survey leads to negative affect reported later. Such experiments will help us understand how dynamic psychological and cognitive constructs like attention and affect unfold over time in ADHD.

Second, the sample covered a relatively wide age range, and was predominantly female. This makes it unclear whether our results generalise equally well across younger and older populations, and between females and males. Future studies, such as those with within-subject longitudinal designs as described above, should consider exploring the hypotheses within specific combinations of age range and gender/sex.

Third, our results may have limited generalisability to the clinical ADHD population, given we did not restrict recruitment to patients with formal diagnoses of ADHD, or include other expert or observer-reports to quantify ADHD symptom levels. Instead, we used a self-report version of ADHD symptom assessment and included formal clinical diagnosis/medication as a participant-reported covariate. Notably, a majority (58 out of 101) of our participants did have a formal diagnosis of ADHD, and their median levels of self-reported inattention and hyperactivity symptoms were 1.6 and 1 standard deviation above general (non-clinical population) symptom levels, suggesting our results are generally most relevant around sub-clinical levels of ADHD symptoms. Importantly, in supporting Supplementary Analyses (reported in Supplemental Tables S10–S13), we found that our primary results were largely invariant to the clinical status (diagnosis/medication) of the participants. This is consistent with growing consensus that behavioural and functional outcomes associated with ADHD covary almost entirely linearly with dimensional symptom levels of ADHD, and researchers have found that dichotomous or non-linear thresholds (like clinical vs. non-clinical categorisation) offer little to no improvements in behavioural or functional outcome prediction, arguing such categorisation should be validated on largely pragmatic grounds where necessary (Arildskov et al., 2024; Marcus & Barry, 2011). Moreover, there is at least some evidence to suggest that there may be considerable day-to-day variability in the symptom levels of all ADHD symptom dimensions (Schmid et al., 2020), making it likely that some participants included in the present sample may have completed EMA reports on days when their objective symptom levels were relatively high and clinically significant. In summary, while we did not implement clinically supervised assessment of ADHD symptoms, or include case-control groups for our analyses based on clinical diagnosis, our sample was composed of a majority clinically diagnosed sample, with median self-reported levels of ADHD symptoms across the sample that are relatively medium-to-high compared to the general population, making our results likely to be most valid and generalisable to a population with sub-clinical symptom levels of ADHD.

Fourth, our analyses did not control for the momentary environmental context within which the participants’ EMA responses were made. It is possible, therefore, that our results may be confounded by such unmeasured environmental covariates or moderators. Indeed, this is an important direction of research, and future studies can explore the effects of such additional contextual variables on the relationship between ADHD symptomatology, mind wandering, and affect. For example, prevalence rates of addiction (such as smoking) remain disproportionately high among individuals with ADHD (Whalen et al., 2002), and there is some evidence to suggest that smoking may momentarily reduce clinical ADHD symptoms (Gehricke et al., 2006). This can set off both feedforward and feedback effects on attention, affect, and substance-use. Future research can investigate how momentary environmental context and activities may interact with ADHD symptoms to longitudinally and causally predict momentary attentional and affective states.

Finally, although EMA is generally considered to be a sufficiently reliable method of assessing momentary cognitive states (Csikszentmihalyi & Larson, 1987; Moskowitz & Young, 2006; Shiffman et al., 2008), it is still a subjective self-reported measure. This is a limitation of the inherently subjective nature of the constructs of mind wandering (especially with respect to intentionality) and affect. More research is needed to explore whether there exist objective and direct measures of momentary cognitive and affective states. Further, future research should investigate how other trait-like correlates of MW, such as depression, or sluggish cognitive tempo (Fredrick et al., 2020), may interact with specific ADHD symptom dimensions in predicting intentional or unintentional MW.

Conclusion

Taken together, our results highlight the importance of differentiating between intentional and unintentional mind wandering in ADHD (Seli, Smallwood, et al., 2015), as well as the different symptom dimensions of ADHD (Nikolas & Burt, 2010; Willcutt et al., 2012), in considering their associations with affective well-being. They also suggest that unintentional MW may not be uniquely associated with all ADHD symptom dimensions (Bozhilova et al., 2018). Instead, our results paint a more complex picture of the relationship between clinical levels of ADHD symptom dimensions, mind wandering, and affective valence.

Author Biographies

Yudhajit Ain is a PhD student in Psychology at the University of Calgary. He has an MSc in Psychology (University of Calgary), and an MS in Biology (Indian Institute of Science Education and Research, Bhopal). His research focuses on cognitive neuropsychology and mental health.

Simrit Rai is an MS student in Community Health Sciences at the University of Calgary. She holds two BScs in Psychology and Cellular, Molecular, Microbial Biology from the University of Calgary.

Ann Galbraith is a Senior Systems Analyst with the City of Calgary. She has a BA in Psychology from the University of Calgary.

Jonas Buerkner is an MS student in Sensors and Cognitive Sciences at the Technische Universität Chemnitz. He has a BA in Psychology from the University of Calgary.

Jessica R. Andrews-Hanna is an Associate Professor of Psychology at the University of Arizona and the director of the Neuroscience of Thought and Emotion lab. In her lab, she studies functional and dysfunctional internally-guided thought and their neural underpinnings, using a number of methodological approaches, including task-related and resting-state functional MRI, psychophysiology, behavioral tasks and smartphone apps (ecological momentary assessment, ambulatory assessment).

Brandy L. Callahan is a Canada Research Chair and an Associate Professor in Psychology at the University of Calgary, and the principal investigator of the Lifespan Brain Health Lab. She is a neuropsychologist interested in understanding brain health and cognition across the lifespan using a wide range of methodological approaches in diverse clinical populations.

Julia W. Y. Kam is an Associate Professor of Psychology at the University of Calgary and the director of the Internal Attention Lab. Her lab research examines the functional and neural basis of thought, including experiences like mind wandering, using a variety of neural, experiential, and naturalistic methods, in healthy and clinical populations.