Abstract
The study of self-control occurs in many different types of experimental settings using a wide range of methodologies. In addition, measures of self-control vary in their procedures and operational definitions from simple questionnaires to complex scenarios where individuals must choose to act or not. The present summary draws on trends within the literature using widely accepted measures of self-control. The measures are organized based on established paradigms in the literature and focus on three categories: executive functioning tasks, delay of gratification tasks, and subjective-report surveys. We also include an “additional measures” category to capture measures that do not readily fit in these three categories. Finally, we discuss recent approaches to the scientific exploration of self-control and integrate the categories of measures used here within these approaches. This integration incorporates a wide range of research paradigms and provides direction for future studies.
Understanding and measuring self-control are common topics in applied psychology research. Self-control, the ability of an individual to override, alter, and interrupt current desires or impulsive behaviors in advancement of a longer-term goal, is integral to everyday decision making, goal achievement, and long-term functioning (Fitzsimons et al., 2015; Fujita, 2011; Mansell & Marken, 2015). Consequently, assessing self-control is an important component in a wide range of research in behavioral and social sciences. Cognitive neuroscience has explored the link between self-control and attention deficits (Berger et al., 2007), developmental psychology has reviewed the strategies children use to bolster self-control (Moffitt et al., 2011; O’Leary & Dubey, 1979), and cognitive ergonomics has explored the reciprocal relationship between decision making and impulsivity (Vohs et al., 2008). Moreover, transportation researchers apply theories of self-control to predict risks while driving (Keane et al., 1993), educators examine the impact of self-control on academic performance (Englert & Bertrams, 2013; Oaten & Cheng, 2006; Tangney et al., 2004), and personality researchers have shown robust links between self-control and addiction (Kim et al., 2008; Schilbach, 2019). Among these studies, self-control appears as an independent, dependent, moderating, and mediating variable, and has been broadly applied to all manner of research questions ranging from dieting to exercise to religion to finances (Iso-Ahola, 2013; Liu et al., 2019; Reynolds & Baumeister, 2016; Sullivan et al., 2015). As such, the scientific literature suggests that self-control is a fundamental element of human behavior.
In part due to the ubiquity of self-control research in different areas within the behavioral and social sciences, self-control has become a challenging construct to define and measure (Evenden, 1999; Vadillo et al., 2016). The construct of self-control is further complicated by a lack of uniform terminology in self-control studies across disciplines, with certain terms referring to multiple concepts (e.g., “self-control” describing both a state and trait, see Milyavskaya et al., 2019) and instances of multiple terms referring to the same concept without meaningful empirical differences (e.g., considering grit, low impulsivity, and conscientiousness separately despite research that places all three concepts in the conscientiousness domain, see Milyavskaya et al., 2019; Roberts et al., 2014). Each research discipline can introduce a new method or modify an existing method of measuring self-control, effectively creating a new operational definition of self-control with unique psychometric properties, and yet maintaining the same basic definition (Duckworth & Kern, 2011). These methods can include itemized surveys, specialized vignettes, cognitive and reaction time tasks, and physiological metrics. When these domain specific variabilities are combined with laboratory time constraints, budgetary limitations, and specialized software and measurement apparatuses, the result is high methodological variability. Albeit, when considering the many ways that self-control is operationalized, it is important to note that there does not seem to be a single way or even a best way of measuring it. Rather, the literature is filled with best practices and limitations to consider when assessing various operationalizations of self-control (see Gillebaart, 2018).
Self-control has traditionally been seen as a combination of effort and inhibition (Gillebaart, 2018). However, limiting self-control to these two areas creates a simplistic view of a complex human behavior. For example, to obtain long-term goals, one must inhibit counterproductive behaviors but also initiate a plan for productive behaviors which can go beyond the literal effort to carry out the behaviors. Perhaps one of the most effective uses of self-control is to initiate and maintain strategies that lead to the long-term goal but with the minimal possible effort instead of simply pushing forward with full effort but little planning. Because there are many different elements to the successful integration of behaviors that lead to a long-term goal, there are many measures of self-control to assess these different components. The purpose of this paper is to provide a review of the major categories of self-control measures and integrate the measures with the current approaches to and debates about the study of self-control.
Categories of Self-Control Measures
Researchers have previously noted the challenges of defining and organizing diverse measures of self-control (Evenden, 1999). To best address this issue, the current summary borrows from the organization established by Duckworth and Kern (2011) who implemented a meta-analytic study on the convergent validity of self-control measures. In addition, our organization reflects the impulsivity categorization of Mackillop and colleagues (2016). Duckworth and Kern identified four common methodological categories of self-control measures: executive functioning tasks, delay of gratification tasks, and self- and informant-report questionnaires. Mackillop and colleagues identified three measures of impulsivity: impulsive actions, impulsive choices, and impulsive personality traits. These categories for impulsivity are operationally similar to Duckworth and Kern’s categories since executive functioning tasks access the ability to inhibit impulsive actions, delay of gratification tasks access the ability to inhibit impulsive choices, and self- and informant-report questionnaires can be used to assess impulsive personality traits. Although Duckworth and Kern structure their meta-analytic findings with self- and informant-report as two separate categories, it is also logical to collapse those into one category that relies on subjective measures. As such, the Duckworth and Kern (2011) and the Mackillop and colleagues (2016) studies propose a three-faceted organization of self-control. For the purposes of the current paper, we will use the three categories: executive functioning tasks, delay of gratification tasks, and subjective-report surveys to organize the self-control measures.
It is important to note that subjective-report surveys can assess more than one construct of self-control. In addition to documenting whether a subjective-report survey is completed by self or an informant, it is also important to distinguish between self-control as a personality trait and self-control as a temporary state of ability (Christian & Ellis, 2011). As such, in this manuscript, self-control questionnaires are not only organized by self- and informant-report, but also by trait and state.
Finally, there are other types of self-control measures that were not captured in the Duckworth and Kern (2011) and Mackillop and colleagues (2016) studies. Advancements in technology and the growth in the popularity of applied psychophysiological metrics have also given rise to the use of physiological indices in self-control research. Psychophysiological measures of stress are intrinsically related to cognitive and physical effort and can be considered as indicators of the effort involved in self-control (e.g., Appelhans & Luecken, 2006). In addition, there are other physical effort measures and mental skills measures that are commonly used to assess self-control. This summary uses a category of “Additional Measures” to refer to these further methods of evaluating self-control.
Executive Functioning Tasks
Executive Functioning Tasks.
Another type of executive functioning task is cognitive persistence tasks such as the anagram task and tracing puzzle task. These tasks require participants to mentally problem solve difficult or impossible puzzles for as long as they are willing (Dzhogleva & Lamberton, 2014; Wallace & Baumeister, 2002). Another type of persistence task is the thought suppression task, which requires participants to effortfully moderate their free thoughts (Job et al., 2013). For thought suppression tasks, the researcher selects a conscious thought that the participant should avoid thinking about, and the participant reports when they have had the thought. The researcher may choose food items as forbidden thoughts when studying diet related self-control, cigarettes when studying addiction, or worries when studying anxiety (Barnes & Tantleff-Dunn, 2010; Erskine et al., 2010; Silva et al., 2016).
Executive functioning performance is often quantified as the number of false positive responses or as time-to-completion with errors resulting in a time penalty. False positives would include the participant responding routinely when a stop-signal was present or when they read the word instead of reporting the color of the word. Requiring more time to respond or making more errors are considered instances of decreased self-control. Because response time is a primary metric, executive functioning tasks are often administered using specialized software packages such as the Laboratory Behavioral Measures of Impulsivity package (Dougherty et al., 2005), or the National Institutes of Health toolbox cognitive function battery (Weintraub et al., 2013), while some tasks such as the Stroop Test may still be administered in paper form. One unique advantage of using reaction-based metrics to measure self-control is their implicit relationship with the self-control construct. Executive function tasks do not rely on a conscious evaluation of self-control like some methodologies, and therefore may be considered more resistant to demand characteristics (Strack & Deutsch, 2004). For instance, participants completing a questionnaire about self-control may be acutely aware of how they wish to be perceived and alter their responses accordingly. However, participants would not generally associate a reaction time task with self-control, allowing for a somewhat covert measure of self-control without the need for deception.
Persistence task performance is often quantified as pass/fail, time-to-failure, or number of attempts. In many instances, failure is the point at which the participant no longer tries to continue, for example, to no longer attempt to solve the puzzle. Choosing to attempt a given task for a shorter duration or making fewer attempts are considered instances of low self-control. During thought suppression, low self-control would be a function of how many times a forbidden thought was reported within a period of time. Actual mental performance on a given task may also be analyzed independently but would not generally reflect a persistence-based measure of self-control. For instance, a task which measures the number of attempts a participant was willing to make to solve a difficult puzzle may also be analyzed based on how well the participants did on the puzzle.
It is important to note that executive functioning tasks require sustained self-control to complete, thus potentially influencing performance on succeeding tasks. For example, the 135-item color-word interference version of the Stroop task has been purposefully used to exhaust individuals (Wallace & Baumeister, 2002). Similarly, thought suppression has been used as a measure of self-control but has also been shown to deplete internal resources and limit performance on subsequent self-control measures (Muraven et al., 1999). In fact, prompting participants with a forbidden word paradoxically evokes forbidden thoughts, so it can be argued that the effect of the prompt is stronger than the effect of the task itself (Tolin et al., 2002).
Another consideration is the impact of fatigue on executive measures of self-control (Pilcher et al., 2007). Sleep-related fatigue has been shown to have a significant negative impact on reaction time, response lapses, and error monitoring (Morris et al., 2015; Tsai et al., 2005). As a result, fatigue can mimic the effect of depleted self-control while masking the true relationship between the two variables (Pilcher et al., 2015). Similarly, age is another factor that can skew performance on executive functioning tasks. When compared to younger adults, older adults (> 54 years) were found to have greater within-person reaction time variability across time for both simple and choice reaction tasks (Hultsch et al., 2002). In this case, systematic variability in choice reaction time performance, as is seen with the flanker task and stop-signal task, may not be entirely due to self-control if an older population is used.
Delay of Gratification Tasks
Delay of Gratification Tasks.
Delay of gratification performance can be quantified relative to a selected reward or a pass/fail on a resistance test. Monetary choice tasks are quantified relative to the value of a selected reward and produce an estimated discount rate that reflects the disparity of that reward from a reward not selected (Kirby & Maraković, 1996). For instance, a participant may choose to receive a reward of $10 now instead of waiting one week for $15, and an estimated discount rate would be calculated using the $5- and seven-day dispersity. The Monetary Choice Questionnaire was developed with a set array of fixed values within the questionnaire items that produce particular discount rates (Hamilton et al., 2015). As an alternative, the computerized Richards Task can be used to dynamically modify a discounting rate and arrive at an indifference point outside of the predefined values of the Monetary Choice Questionnaire (Richards et al., 1999). Versions of a consumable choice reward task using food or drink can also be analyzed for discounting rate, although they tend to be valued less than money rewards (Charlton & Fantino, 2008). More commonly, performance using consumable rewards is assessed by simply noting whether the participant was able to resist the smaller reward sooner in favor of the larger reward later (Drobetz et al., 2014).
It should be noted that if actual rewards are used during a monetary choice task, procedures can be prohibitively expensive for larger studies. However, several studies have reported equivalent results between hypothetical and actual rewards, so actual rewards may not be necessary (Bickel et al., 2009; Matusiewicz et al., 2013). Similar to the executive functioning tasks, performing a delay of gratification task also seems to limit performance on succeeding measures of self-control such as measures that require mental calculations (Kirby et al., 1999; Vohs et al., 2008).
Subjective-Report Surveys
Subjective-Report Surveys.
There are two primary factors that may be used to categorize subjective-report surveys. First is whether the scale was designed as a measure of trait self-control or state self-control. Trait scales measure self-control as a long-term personality trait and include items such as “are you generally calm,” as is seen in the classic Eysenck Impulsivity Scale (Eysenck & Eysenck, 1978). State scales, on the other hand, measure transient and momentary feelings of self-control, and include items such as “I feel calm,” as is seen in the Sate Self-Control Capacity Scale (Christian & Ellis, 2011). In addition, self-control measures can provide additional information beyond what traditional personality measures might indicate (Bazzy et al., 2017). For example, combining self-control with a personality measure of conscientiousness provides a reliable prediction of many life outcomes (Duckworth & Seligman, 2017).
Some measures of personality can be used as trait measures of self-control even if they are not traditionally viewed as self-control measures. Personality measures of conscientiousness, for example, seem to be good predictors of self-control, and have even been considered as a subcategory of self-control measures (Moffitt et al., 2011). Additionally, the Big Five Conscientiousness scale is negatively correlated with bouts of anger, a common item assessed in self-control subjective-report surveys (e.g., The Interview for Antisocial Behavior, The Eysenck Impulsiveness Scale) (Eysenck et al., 1984; Jensen-Campbell et al., 2007; Kazdin & Esveldt-Dawson, 1986). Another personality-like trait that relates to self-control is the grit construct. The concept of grit shares many of the same elements as self-control and includes the personal progression or drive towards a goal. Subjective measures of grit can be considered a long-term trait self-control measure and have been found to correlate highly with other measures of self-control (Duckworth et al., 2007). Indeed, some researchers consider conscientiousness as the true underlying construct of grit (Rimfeld et al., 2016).
The second primary factor that can be used to categorize surveys assessing self-control is whether the scale is completed as a self-report questionnaire or an informant-report questionnaire. Self-report surveys require the participant to report their own perception of their self-control and include items such as “I have trouble concentrating,” as is seen in the Brief Self-Control Scale (Tangney et al., 2004). Informant-report surveys require the participant to report their perception of another individual’s self-control and include items such as “the individual can concentrate on one thing at a time,” as is seen in the Self-Control Rating Scale (Kendall & Wilcox, 1979). Informant-report questionnaires are often used for child participants, given their limited capacity for responding to complex questionnaires (Rorhbeck et al., 1991).
Subjective reports are often quantified according to a predetermined formula or summation score and designed with reverse coding. Many surveys are comprised of subscales and are analyzed holistically and divisionally. Such is the case with the UPPS Impulsive Behavior Scale, which originally measured self-control through the subscales of urgency (the tendency to engage in impulsive behavior when experiencing a negative affect), premeditation, perseverance, and sensation-seeking behavior (Whiteside et al., 2005). The UPPS Impulsive Behavior Scale has subsequently been modified to include a positive urgency (the tendency to engage in impulsive behavior when experiencing a positive affect, see Cyders & Smith, 2008) subscale (the UPPS-P, see Lynam et al., 2006) and abbreviated for quicker administration (the SUPPS-P, see Lynam, 2013). In general, subjective-report surveys are associated with an ease of administration, ease of analysis, and shorter administration time (Porter, 2004). Each of the surveys listed in this summary can be administered in a paper or digital format.
As mentioned previously, subjective-report surveys may make the participant aware of the objective of the task or study. Many surveys ask questions such as “I’m good at resisting temptation,” which may then influence participant demand characteristics on a following task (Tangney et al., 2004). Research assessing willpower suggests that simply becoming aware of self-control can alter performance (Job et al., 2010). It is also worth noting that trait self-control surveys may be poor predictors of transient self-control. Many models of self-control assume a limited capacity and predict short-term modulation, yet many trait scale items assess long-term behavioral patterns suggesting that self-control is a complex behavior that can be viewed as a multi-construct phenomenon.
Finally, although we have focused on three methodological categories of measuring self-control, it can be argued that some tasks could qualify for multiple categories. For instance, the Monetary Choice Questionnaire is a delay of gratification task but could also be seen as a state self-report scale because of its questionnaire format (Kirby, 2009). In addition, many tasks in one category may share similarities with another category. In many respects, persistence-type tasks are also delayed gratification tasks because the participant delays the catharsis of quitting when they persist (Bembenutty & Karabenick, 2004). These considerations make it clear that self-control is a complex combination of thoughts, choices, and actions that can be measured in a variety of ways that likely overlap to some extent.
Additional Measures
Other Measures.
Unlike the previous methods covered here, physiological indices are quantified according to the physiological response rather than task performance or conscious report. This implies that physiological indices are not a direct measure of self-control, as are the other methodological categories, but a measure of whether self-control is being used. It is important to note that psychophysiological variability is dynamic, transient, and seldom reflects a single mechanism, making it difficult to attribute a physiological change to a single psychological construct. Cardiac activity, for example, has been shown to vary with self-regulatory effort (Ingjaldsson et al., 2003), but cardiac variability is also a function of affect, cognitive workload, physical work, and the underlying homeostatic balance (Appelhans & Luecken, 2006; Hjortskov et al., 2004).
There are also physical tasks that have been used as measures of self-control (see lower portion of Table 4). One common example of a physical task is the handgrip task, which requires participants to squeeze a hand-held apparatus for as long as they are able (Fujita et al., 2006; Leahey et al., 2014). The handgrip apparatus may require an isotonic or isometric contraction and, in addition to duration of grip force, may record amount of force applied. Sustained acts of physical strain may also include tasks such as breath holding, but those are less commonly seen in the literature and have more variable outcomes (Hajek et al., 1987; Muraven et al., 1998).
Another measure that is associated with self-control is mindfulness. Mindfulness is often measured through self-report surveys that can be used in conjunction with task measures or subjective measures of self-control (see lower portion of Table 4). One commonly used measure of mindfulness is the Five Facet Mindfulness Questionnaire (Baer et al., 2006), a 39-item measure of trait mindfulness. The questionnaire contains items such as “I perceive my feelings and emotions without having to react to them” and “I’m good at finding words to describe my feelings.” A further subjective measure of mindfulness is the Mindfulness Attention Awareness Scale (Brown & Ryan, 2003), a 15-item measure of trait mindfulness. The survey contains items such as “I rush through activities without being really attentive to them” and “I find myself listening to someone with one ear, doing something else at the same time.”
Mindfulness is positively related to self-control (Chiesa et al., 2011). More specifically, increased mindfulness is related to improved emotional regulation (MacDonald, 2021; MacDonald & Baxter, 2017), increased delay of gratification (Lawler et al., 2019; MacDonald, 2021), and decreased impulsivity (Fetterman et al., 2010). Similar to the subjective reports used to quantify self-control, mindfulness surveys are often scored using a predetermined formula or as a summation score. They can consist of subscales as with the five components of the Five Facet Mindfulness Questionnaire and can be analyzed holistically and by individual components.
Integration and Future Directions
A number of research studies have addressed different approaches to how to define and interpret self-control. One recent study examined the structure of self-control by exploring the relationships among the different types of tasks and surveys that purport to measure self-control (Eisenberg, et al., 2019). The study concludes that although tasks and surveys are designed to measure a common construct of self-control, they are not related to each other. Eisenberg and colleagues also examined how well the tasks and surveys predict real-world outcomes (e.g., socioeconomic outcomes, physical and mental health) and found that surveys modestly predict real-world outcomes, but the tasks do not. Although our current study cannot directly address these results, researchers could consider how to best utilize the tasks covered here for future studies. Based on the results from Eisenberg and colleagues, it seems particularly important to choose a combination of self-control tasks and surveys to better ensure that the results reflect the broad construct of self-control. Future research is needed to better understand how different self-control measures can be used and combined to assess such a complex human behavior.
Another study specifically focused on the concept of self-regulation versus self-control (Gillebaart, 2018). This paper concluded that self-regulation can be viewed as the broader effort of managing goal pursuit including setting standards, monitoring progress, and monitoring any discrepancies that may occur while self-control can be viewed as the actual skills and behaviors that persons use to make progress toward the goal. The tasks described in the current paper can be employed to effectively measure the types of skills and behaviors needed as part of this self-control effort. It is important to note, however, that the measures described in the current paper do not match well with this broad definition of self-regulation. Future research could establish measures that assess the components of self-regulation described by Gillebaart. In addition, future research could focus on measures that assess the broad effort of goal pursuit and goal management to better operationalize Gillebaart’s suggested approach to self-regulation and self-control.
An earlier approach to self-control specifically focused on desire-goal conflict as a primary activator of self-control (Kotabe & Hofmann, 2015). The strength of the desire, level of skill or ability to address the desire, and the power of potential competing goals all impact the amount of self-control effort that one would generate to reach a goal. In this approach an “exertion cluster” is identified that focuses on the self-control capacity, motivation, and effort needed to achieve a goal. Most of the self-control tasks and scales described in the current study can fit as a measure of some part of this exertion cluster. In fact, the delay of gratification tasks could be particularly useful for exploring desire-goal conflicts. Future research could establish measures that better reflect the other aspects of their theory including what they call the “activation cluster” which includes components like desire and desire-goal conflict and their “enactment constraints” which integrates other areas that could impact the person’s ability to reach their goal.
Another approach to conceptualizing self-control has been presented by Fujita (2011) and specifically focuses on a dual-motive perspective of self-control. In this model, self-control includes inhibition of impulses but also integrates other ways that people can promote their immediate self-control effort to reach a longer-term or more distal goal. The basic Fujita model is the dual motive of proximal motives interfering with distal goals. This perspective is similar to the desire-goal conflict described by Kotabe and Hofmann (2015). In both models, the individual must exert self-control to overcome immediate desires to reach a longer-term and often more abstract goal. The groups of self-control measures summarized in the current study can be used to assess self-control capacity or skill to overcome immediate desires in an effort to reach a goal. Future studies can contribute to this effort by examining how specific self-control tasks and surveys measure dual motives that are a common component of self-control.
An additional psychological construct that could impact an individual’s ability to manage their self-control is willpower. Some researchers suggest that subjective measures of willpower may be an effective way of predicting self-control. For example, a simple survey on willpower beliefs was used to predict behaviors that are associated with diminished self-control (Job et al., 2010). This relationship suggests that measures of personal beliefs about willpower may be a critical element of self-control (Wan & Sternthal, 2008) and that internal motivation may be part of the larger construct of self-control effort. Future studies can be designed to address how willpower may fit in as part of the self-control effort described in the different models above.
Each of these approaches adds to and helps define how we can best conceptualize and measure self-control. These approaches, however, cannot elaborate on all potential moderators of a complex human behavior like self-control. For example, factors such as social support may moderate the effectiveness of self-control and motivation (Pilcher & Bryant, 2016). Future research is needed that focuses on this and other potential moderators to help provide information that can be used to more clearly understand self-control across a range of real-world situations. In addition, future research could focus on different types of self-control strategies that different individuals may develop and how to best measure those strategies. This is an area where the current measures of self-control fall short in terms of documenting self-control effort in scenarios that better reflect real-world conditions. This is particularly important given Eisenberg and colleagues’ (2019) findings that self-control surveys predict real-world outcomes better than self-control tasks.
Conclusions
To date, there does not seem to be a single way or even best way to measure self-control. Some studies stress the importance of developing a task that will best reflect the applied scenario where it is being used (Benoit et al., 1996), while other studies focus on self-control as a basic construct best understood using cognitive neuroscience and laboratory-based measures (Tabibnia et al., 2011). Many studies in the current summary use a variety of tasks or surveys to assess self-control such as decision-making tasks, attentional tasks, personality trait surveys, or some combination of the three, and still place the findings under the umbrella term of self-control.
Self-control is an important psychological construct that has been quantified using several operational definitions. Researchers have developed a number of unique approaches for assessing self-control, each with their own advantages, disadvantages, and relationship within the prevailing models of self-control. Although some may see this operational diversity as a shortcoming, the flexibility of the self-control construct highlights the ubiquity of and importance of the observed effects. The current summary sought to synthesize the components of common methodologies and integrate them with recent theoretical discussions of self-control. The present paper can help advance the field by providing a means for future research to focus on diverse self-control strategies and the existing measures. In addition, this paper can lead to the development of new measures to better quantify the aspects of self-control discussed in some of the theoretical approaches included here as well as provide a useful description of the current types of self-control measures.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
