MEASURING INDIVIDUAL DIFFERENCES IN EMOTION REGULATION: THE EMOTION REGULATION PROFILE-REVISED (ERP-R)

The main purpose of this study was to validate a new instrument aimed to assess emotion regulation: the Emotion Regulation Profile-Revised (ERP-R). Exploratory factor analyses yielded two theoretically meaningful factors: down-regulation of negative emotions and up-regulation of positive emotions. Internal reliability scores of the two factors were good. Findings showed evidence of convergent/discriminant validity, with ERP-R scores being independent of non verbal reasoning and verbal skills while positively related to emotional intelligence and to relevant personality dimensions. There was also preliminary evidence of criterion validity. ERP-R scores also demonstrated incremental validity to predict a number of criteria over and above emotional intelligence and emotional stability. Overall, the results show a clear 2 factors solution for the ERP-R and high correlations with convergent and divergent scales as well as good criterion and incremental validities.


Introduction
As more and more evidence suggests, emotions do not only colour people's lives, but are absolutely essential to their survival and adaptation (Cosmides & Tooby, 2000).For instance, emotions facilitate the detection of threatening stimuli (e.g., Öhman, Flykt, & Esteves, 2001), prepare the organism for specific behavioural responses (Frijda, 1986), enhance memory for significant events (Luminet & Curci, 2009;Phelps, 2006), increase the speed and accuracy of decision-making processes (see Bechara & Damasio, 2005), and guide social interactions (Keltner & Kring, 1998).Although emotions are very efficient systems, they are nevertheless very old devices.While emotions were fully adjusted to the life of our Pleistocene ancestors, they are not quite adapted to the life of modern humans (Gross, 2007;Mikolajczak, 2009).For instance, getting nervous and angry in traffic jams is totally useless.According to Gross and Thompson (2007), emotions become dysfunctional when they are of the wrong type, when they come at the wrong time, or when they occur at the wrong intensity level.In these cases, emotions must be regulated.The two most common circumstances in which people regulate their emotions are (1) when their emotions impede goal achievement and (2) when their emotions do not match with the group's emotional display rules.Emotional regulation (ER) refers to the processes through which individuals influence which emotions they have, when they have them, and how they experience or express these emotions (Gross, 1998).ER includes a wide range of automatic and controlled physiological, behavioural, and cognitive processes (Gross, 2001).
Though emotion regulation in everyday life predominantly involves the down-regulation of negative emotions, most individuals also attempt to regulate their positive emotions (Gross, Richards, & John, 2006).Positive emotions can be (1) down-regulated (e.g., when we try to decrease love for a colleague who is married), ( 2) maintained (e.g., when we engage in social sharing in order to prolong the effects of a positive event), and (3) up-regulated, such as when we try to enjoy a long planned vacation despite disappointing weather, food, and housing (Mikolajczak, 2009).
Despite the paucity of research on the topic, the ability to maintain and upregulate positive emotions is of particular relevance for well-being and human flourishing (Tugade & Fredrickson, 2007).Indeed, positive emotions broaden the scopes of attention and cognition, thereby facilitating the building of personal resources and initiating upward spirals toward increasing emotional well-being (Fredrickson, 2001;Fredrickson & Branigan, 2005).Numerous studies show that positive affect engenders success across multiple life domains, including marriage, friendship, income, work performance, and health (for a review see Lyubomirsky, King, & Diener, 2005).
The above-mentioned literature leads to three major points: (1) emotion regulation has a major impact on the most important domains of life, (2) there psycho.belg.2011_1.bookPage 50 Tuesday, December 7, 2010 8:34 AM has been an increasing interest in ER research, with the number of studies growing exponentially each year (see Gross, 2007), and (3) there is a lack of instruments measuring regulation of positive emotions and individual differences in ER.The need to be able to measure individual differences in this field could be interesting for both research and clinical fields.For instance, from a clinical point of view, it is important to know which ER strategy is employed by an individual to establish his/her ER profile.The development of the Emotion Regulation Profile-Revised (ERP-R) is based on these three elements.The ERP-R is a revision and an extension of a previous unpublished version called Emotional Regulation Profile-Questionnaire (ERP-R; Quoidbach, Nelis, Mikolajczak, & Hansenne, 2007) that did not include the regulation of positive emotions.
The ERP-R (see Appendix) is a vignette-based measure developed in French.It comprises 15 scenarii describing different types of emotion-eliciting situations.Each scenario features a specific emotion (e.g., anger, sadness, fear, jealousy, shame, guilt, joy, contentment, awe, gratitude, pride) and is followed by eight possible reactions: four considered as adaptive in the literature and four viewed as maladaptive (see below for a detailed presentation of the strategies).Respondents are required to select, for each scenario, the strategy(ies) that best describe their most likely reactions in the situation.
Compared with existing ER instruments, the ERP-R presents several advantages.First, most ER measures provide a very general idea of an individual's level of ER competence (they provide a global ER score) but fail to indicate which strategies people use to achieve those scores.This problem is well represented in Emotional Intelligence tests or ER-related dimensions of personality tests (i.e., neuroticism).Those tests comprise ER as a core dimension but they provide only a very general idea of one's ER competence.This may be sufficient in selection settings but might not be detailed enough in clinical settings where therapists are often interested in determining which strategies people use to (mis)regulate their emotions.The ERP-R, on the other hand, provides a detailed emotion regulation profile of the individual.Namely, the ERP-R not only provides information about how a person regulates his/her emotions, but it also highlights the regulation strategies used.Second, the items of most existing ER measures are very general ("I am usually able to regulate my emotions"), and are therefore difficult to answer for some people who have to aggregate their emotion regulation abilities across a wide range of situations.This is the reason why the ERP-R is a vignettebased instrument.Third, contrary to the two existing ER vignette-based measures (i.e., Freudenthaler & Neubauer's Emotion Management Abilities test, 2007;Mayer, Salovey & Caruso's Emotional Intelligence Test, 2002) in which the functional-dysfunctional nature of the strategies are based on consensus, the functional-dysfunctional nature of the ERP-R strategies is deter-psycho.belg.2011_1.bookPage 51 Tuesday, December 7, 2010 8:34 AM mined according to the empirical evidence available in the literature.In the consensus scoring method, the correct answer is the answer that has been most frequently chosen.If among four responses, 70% of individuals choose the response "a", any respondent who chooses "a" will be credited by 0.70 point.If the respondent selects an answer that has been chosen by only 45% of people, s/he will be credited by 0.45 point.This method of scoring is highly problematic when it is applied to complex problems for which, by definition, only a small number of individuals can give the correct answer.Tests based on consensus scoring can thus hardly discriminate between "ER geniuses" and "ER dumbies".Moreover, research has shown that popular beliefs cannot be relied upon to identify effective strategies (see Loewenstein, 2007).Accordingly, strategies presented in the ERP-R have been chosen on an empirical basis.Functional strategies were deemed as such because they were repeatedly found to be associated with a decrease of physiological activation in experimental studies and/or with positive indicators of mental/somatic health in clinical studies.Dysfunctional strategies were deemed as such because they were repeatedly found to be associated with an increase of physiological activation in experimental studies and/or with negative indicators of mental/somatic health in clinical studies.Finally, most ER instruments target only the down-regulation of negative emotions except the MSCEIT (the Mayer-Salovey-Caruso Emotional Intelligence Test; Mayer et al., 2002) and the EMA (the Emotion Management Abilities; Freudenthaler & Neubauer, 2007).This is surprising as there are four forms of ER: down-regulation of negative emotions, down-regulation of positive emotions, up-regulation of negative emotions, and up-regulation of positive emotions (Gross, 2007).As shown by Gross et al. (2006) the first and the last forms are by far the most frequent ones.Among these, the down-regulation of negative emotions is the most common.The second most frequent form of emotional regulation relates to the maintenance or increase of positive emotions.Given that both the down-regulation of negative emotions and the maintenance or up-regulation of positive emotions predict adaptation, ER measures should at least comprise these two dimensions.Thus, ERP-R evaluates these two forms of regulation.
Nine scenarii evaluate the down-regulation of negative emotions and six measure the up-regulation of positive emotions.We have created two scenarii for the three primary and basic negative emotions (anger, sadness, and fear) and one for the secondary emotions (shame, guilt, and jealousy).The positive scenarii feature six main positive emotions: joy, excitation, pride, gratitude, contentment, and awe.For each scenario, eight reactions are proposed: four adaptive and four maladaptive.
For the negative scenarii, eight distinct regulation strategies are proposed.Those strategies were selected based on literature review on emotion regulation from 1995 to 2008.The examination of these different strategies leads us psycho.belg.2011_1.bookPage 52 Tuesday, December 7, 2010 8:34 AM to aggregate similar ones, resulting in eight broad categories of emotion regulation strategies (four functional, four dysfunctional).The functional strategies are situation modification, attention reorientation, positive reappraisal, and emotion expression.The dysfunctional strategies are learned helplessness, substance abuse, rumination, and acting out.These strategies are detailed hereafter.Even if some of these strategies could be considered as symptoms of various mental disorders (e.g., mental rumination), the core characteristic is the ability to use them in an acute emotional context to improve (or not) well-being.The idea is not to solve a chronic problem.Functional strategies when used in difficult circumstances can help to get back on one's feet and thus are beneficial in maintaining mental health, physical health, quality of social relationships and job performance.By contrast, dysfunctional strategies are those which in difficult situations create a negative spiral and can damage mental health, physical health, quality of social relationships and job performance.None of the dysfunctional strategies represent a disorder in itself, even if they may all lead to such disorders in the long run.For instance, repeated rumination can lead to depression, alcohol abuse can lead to alcoholism, repeated acting out can lead to antisocial behaviour disorder, etc.The reason we selected these strategies rather than other ones is that they are highly predictive of important outcomes in terms of mental health, physical health, and social functioning.Note that these strategies are mediators between the "difficult circumstances" and the outcomes, not outcomes in themselves.For example, repeated rumination associated with other depressive symptoms can lead to depression, but using rumination as an ER strategy to cope with a current situation cannot lead to depression.A person who ruminates but that also uses functional strategies or benefits from a good social support will not necessarily fall into depression.Rumination is therefore an element of vulnerability, but not the pathology in itself.
Situation modification encompasses the strategies aimed at modifying the situation so as to change its emotional load (Folkman & Lazarus, 1980;Gross, 1998;Gross & John, 2003).There are two kinds of such strategies: direct methods, which involve taking some practical actions that influence the situation directly (e.g., fixing the broken car; rehearsing one's talk), and indirect methods in which the intervention of a third person is required (e.g., asking a colleague for some help as a means of finishing a report by the deadline, or asking a stranger to put out his cigarette in order to avoid second-hand smoke; see Mikolajczak, 2009).Taking steps in order to modify an emotioneliciting situation is associated with increased well-being (Billings & Moos, 1981) and better health outcomes (see Penley, Tomaka, & Wiebe, 2002 for a meta-analysis).Moreover, this strategy has been associated with better work performance, both in academic (Struthers, Perry, & Menec, 2000) and organisational (Lee, Ashford, & Jamieson, 1993)  Attention reorientation involves altering how we feel by modifying our attentional focus (Gross, 1998;Nolen-Hoeksema & Morrow, 1993).Attention reorientation may be internal (e.g., thinking about a happy memory, imagining the upcoming holidays) or external (engaging in some pleasurable activity such as listening to music, surfing the web, etc.) (Mikolajczak, 2009).Attention reorientation has been found efficient to decrease negative emotions in emotional situations (e.g., Nolen-Hoeksema & Morrow, 1993;Trask & Sigmon, 1999).
Positive reappraisal involves changing the way we think about the situation (e.g., searching for the silver lining) so as to decrease its emotional impact (Gross, 2001).In the short term, the use of positive reappraisal decreases the subjective intensity of negative emotion (e.g., Gross, 1998).In the long-term, the use of this strategy has been associated with positive outcomes in terms of affective (e.g., Carver, Scheier, & Weintraub, 1989) and social functioning (Gross & John, 2003).Reappraisal has also been found to predict better academic performance (Leroy & Grégoire, 2007).
Emotion expression involves sharing one's emotions with others (Rimé, 2007).Note that the reason why sharing one's emotions is beneficial is not attributable to any catharsis effect (i.e., getting it off one's chest) -as it has long been thought -but to several indirect effects such as the construction or reinforcement of social bonds, the transference of affection and warmth, the expression of esteem, the assistance received in situation modification, and the aid in cognitive reappraisal and in attention reorientation (see Rimé, 2007 for a review).Emotion expression has been associated with improved adjustment in terms of both psychological and physical health (Berry & Pennebaker, 1993;Stanton, Danoff-Burg, Cameron, Bishop, Collins, Kirk et al., 2000).
Learned helplessness involves a passive behaviour accompanied by a feeling of powerlessness.Individuals believe that they are unable to do anything in order to deal with the negative event and consequently do not take any steps to modify it.This strategy is positively related to depression (e.g., Abramson, Seligman, & Teasdale, 1978;Mikulincer, Glaubman, Ben-Artzi, & Grossman, 1991).
Rumination refers to focusing on the feelings and thoughts associated with negative events (Garnefski, Kraaij, & Spinhoven, 2001).Rumination increases the duration and intensity of negative emotions (Bushman, 2002;Morrow & Nolen-Hoeksema, 1990) and predicts the onset, number, and duration of depressive episodes over a 2.5 years follow-up of initially nondepressed individuals (Robinson & Alloy, 2003).Rumination has also been found to impair task performance (Watkins & Brown, 2002).
Substance abuse involves abusing alcohol, anxiolytics, or drugs (e.g., marijuana) in order to avoid or escape an adverse event or its emotional con-psycho.belg.2011_1.bookPage 54 Tuesday, December 7, 2010 8:34 AM sequences.This technique will suppress negative thoughts and emotions temporarily.In addition, this strategy has an impact on the physiological aspect by allowing the body to relax.Although moderate alcohol consumption can have health benefits (for a review, see Baum-Baicker, 1985), alcohol used as a regulation strategy (and thus abused) is associated with poor outcomes in terms of mental and physical health (e.g., Single, Rehm, Robson, & Van Truong, 2000).Anxiolytics can affect alertness, judgment, motor skills, and other cognitive abilities (Aronoff, Erdil, & Hartenbaum, 2005).Marijuana enhances well-being in the short term but at the cost of a drop in cognitive and motor efficiency (see Hall & Solowij, 1998 for a review).
Acting out constitutes an attempt to reduce the emotion by giving way to the action tendency dictated by the emotion: aggression in the case of anger, withdrawal in the case of shame, etc.This strategy is deleterious when it comes to emotion regulation.For instance, physical or verbal aggression leads to exaggerated cardiovascular reactivity in response to provocative stressors (Suls & Wan, 1993), and increases the possibility of developing coronary-heart disease (see Miller, Smith, Turner, Guijaro, & Hallet, 1996 for a meta-analysis).
Regarding positive scenarii, we reviewed the positive emotion regulation literature between 1995 and 2008.This investigation yields different regulation strategies that we aggregated -similarly to negative ones -into eight broad categories: four adaptive (i.e., behavioural display of positive emotions, mindfully savouring the moment, capitalisation, and positive mental time travel) and four maladaptive (i.e., inhibition of emotion expression, fault finding, inattention, and external attribution/nostalgia).
Behavioural display refers to fully expressing one's positive emotions with non verbal behaviours.Evidence for the efficiency of this strategy is provided by many studies showing that the facial expression of emotion may play a causal role in the subjective experience of emotion (Adelmann & Zajonc, 1989).For example, Strack, Martin, and Stepper (1988) found that subjects holding a pen in their mouth in ways that either inhibited or facilitated the muscles typically associated with smiling reported more intense pleasure under facilitating conditions than under inhibiting conditions.Similarly, a pilot trial showed that inhibiting the expression of facial frowning (commonly associated with depression) through Botox injections diminished depressive symptoms (Finzi & Wasserman, 2006).
Capitalising refers to communicating and celebrating positive events with other people (Langston, 1994).Studies have shown that capitalising is associated with increased daily positive affect and well-being, above and beyond the impact of the positive event itself and other daily events (Gable, Reis, Impett, & Asher, 2004;Langston, 1994).Moreover, the wider the net of sharing, the greater the benefits reaped (Gable, et al., 2004).Finally, the expression of positive emotions has also been related to health.Labott, Ahleman, Wolever, and Martin (1990) found that when participants watched a happy video their immune system showed increased activity, but only when they had been instructed to express their emotions.
Positive Mental Time Travel refers to engaging in vivid positive reminiscence or anticipation of positive events (see e.g., Quoidbach, Hansenne, & Mottet, 2008;Suddendorf & Corballis, 2007).Indeed, both correlational and experimental studies have showed that the frequency of positive reminiscing predicts happiness and well-being (Bryant, Smart, & King, 2005;Havighurst & Glasser, 1972;Lyubomirsky, Sousa, & Dickerhoof, 2006).On the other hand, imagining future positive events has also be found to be related to numerous positive outcomes such as subjective well-being and social network size (MacLeod & Conway, 2005), and reduced depressive symptoms (e.g., MacLeod & Salaminiou, 2001).In a recent study, Quoidbach, Wood, and Hansenne (2009) reported that participants who engaged in positive anticipation of future personal events daily for two weeks using vivid cognitive imagery reported a significant increase in happiness over 15 days.
Unfortunately, individuals also engage in maladaptive up-regulation strategies (Feldman, Joormann, & Johnson, 2008;Wood, Heimpel, & Michela, 2003).Based on the above mentioned literature as well as qualitative interviews, the ERP-R distinguishes four maladaptive strategies, which to a large extent are the opposite of the positive ones.
Inhibition of emotion expression refers to the tendency to suppress one's positive emotions for diverse reasons such as shyness, modesty, or fear.Gross and Levenson (1997) showed that the expressive suppression of positive emotions bears physiological costs and leads to a decrease in subjective positive experience (while the suppression of negative emotions also bears costs but does not lead to a decrease in the corresponding subjective negative experience).Gross and John (2003) further showed that the tendency to suppress psycho.belg.2011_1.bookPage 56 Tuesday, December 7, 2010 8:34 AM negative and positive emotions alike leads to less trait positive affectivity and more trait negative affectivity, poorer social support, more depression, and lower life satisfaction and psychological well-being.
Inattention refers to the tendency to engage in activities and thoughts unrelated or detrimental to the ongoing positive event (e.g., worries, uncompleted tasks).Whether during a positive or a negative event, the tendency to worry consistently has the same deleterious effects: it increases anxious and depressive affect (Borkovec, Alcaine, & Behar, 2004), and is associated with increased cardiovascular, neuroendocrine and neurovisceral activity (Brosschot, Gerin, & Thayer, 2006).
Fault finding can also be opposed to a certain extent to savouring the present moment.This strategy relates to a maladaptive focus of attention towards what could have been better or to negative elements in positive situations.Aside from "pure" negative thinking, which is known to be an important characteristic in depression (e.g., Teasdale, 1983), the simple desire to maximise situations has been found to correlate negatively with happiness, optimism, self-esteem, and life satisfaction (Schwartz, Ward, Monterosso, Lyubomirsky, White, & Lehman, 2002).
Negative Mental Time travel encompasses negative reminiscence such as reflecting on the causes of a positive event with an emphasis on external attribution (e.g., "I got an A because the exam was really easy") and negative anticipation of its future consequences (e.g., "These positive feelings won't last").This strategy has been associated with lower self-esteem, greater rumination, and greater depressive symptoms (Feldman et al., 2008).Moreover, external attributions of success have been extensively found to be associated with depression (see Sweeney, Anderson, & Bailey, 1986 for a meta-analysis).
Respondents are allowed to select as many reactions as they want as long as they accurately reflect their typical behaviour in the kind of situation described.This instruction is formulated to avoid a forced-choice between different strategies since individuals can usually use several strategies.Respondents are credited by 1 point per functional strategy selected and by -1 point per dysfunctional one.The ERP-R provides a total score and one score for each of the two factors (down-regulation of negative emotions and up-regulation of positive emotions) as well as specific scores representing the use of each regulation strategy.

Overview
The present study attempts to reduce the limitations of existing ER measures.The aim was the development and the validation of an ER measure that assesses the effectiveness of an individual's typical behaviour in emotional psycho.belg.2011_1.bookPage 57 Tuesday, December 7, 2010 8:34 AM situations more directly than existing measures that tend to assess the individual's knowledge of how to behave, and not their actual emotional management ability per se.First and foremost, we will look at the internal consistency and factor structure of the ERP-R.Next, we will focus on the ERP-R discriminant and convergent validities with regard to indicators of cognitive ability (i.e., non verbal reasoning and verbal skills), trait emotional intelligence, the five-factor model of personality, and cognitive emotion regulation.The liability of ERP-R responses to social desirability will also be examined.Subsequently, we will assess the criterion validity of the ERP-R regarding state affectivity, mental and physical health, propensity to experience various discrete emotions, tendency to experience worry, social support, job satisfaction, and performance.Finally, the incremental validity of the ERP-R over and above emotional intelligence and the big five personality factors will be investigated.

Participants and procedure
Four hundred eighty-one students (360 woman and 121 men) voluntarily completed questionnaires during a Psychology class.The mean age for the sample was 19.5 years (SD = 2.36 years).The whole sample completed the ERP-R along with one or several other measures (during the same or separate sessions).All participants live in Belgium.Participants were recruited to include only French native speakers.The present data came from six samples, and included students in their first two years of university.Participants completed the questionnaires on a paper and pencil form.Responses were anonymous as participants identified themselves with codes.To test if there is a relationship between the ERP-R and work, an additional sample of twenty managers (18 males and 2 females; mean age: 42.40 ± 8.37 years) completed the ERP-R and a measure of job performance.It was not possible for this sample to complete other measures given the length of the testing.The results below involve either one or several samples (depending on the relationships under investigation).The sample size for each measure is indicated in Table 3.

Measures
Emotion Regulation was appraised through the ERP-R described in the introduction (see Appendix).
Non verbal reasoning was evaluated by means of Raven's Advanced Progressive Matrices Test (Raven, 1976), which is one of the most robust meas-psycho.belg.2011_1.bookPage 58 Tuesday, December 7, 2010 8:34 AM ures of general cognitive ability (i.e., supra-ordinate factor "g"; Spearman, 1927).This test consists of 36 problem-series and is independent from language and formal schooling.Each problem consists of 9 figures (arranged as a square) with a missing piece.Below the problem are eight alternative pieces to complete the figure, only one of which is correct.Each set involves a different principle for obtaining the missing piece, and problems are roughly arranged in increasing order of difficulty.
Verbal skills were evaluated using the Mill Hill vocabulary test (Raven, 1943;French translation: Deltour, 1993).The subject must choose among six words which one is the synonym of a target.It comprises 32 target items.Items are arranged in increasing order of difficulty.
Trait emotional intelligence was measured through the French version of the Trait Emotional Intelligence Questionnaire (TEIQue; Petrides & Furnham, 2003; for the psychometric properties of the French adaptation, see Mikolajczak, Luminet, Leroy, & Roy, 2007).The TEIQue (α = .94)consists of 153 items responded to on a 7-point scale (from "strongly agree" to "strongly disagree").It provides scores for 15 subscales, four factors (wellbeing, self-control, emotionality, and sociability) and global trait EI.The 15 subscales of the TEIQue show good internal consistency with scale reliabilities ranging from .70 (self-esteem) to .91 (well-being).
Personality was appraised through the "Description in Five Dimensions" system (D5D, Rolland & Mogenet, 2001), which is a widely used French personality inventory based on the Five Factor Model (FFM; Costa & McCrae, 1992a).This questionnaire assesses the big five dimensions of emotional stability, introversion, openness, conscientiousness, and agreeableness through 55 adjectives (e.g., nervous, reserved, cultivated, compassionate, tidy) rated along a 6-point scale (-3 = does not describe me at all, +3 = describes me perfectly).The D5D factors show good internal consistency with scale reliabilities ranging from .71 (Openness to Experience) to .84 (Conscientiousness).
Cognitive emotion regulation was assessed using the Cognitive Emotion Regulation Questionnaire (CERQ; Garnefski et al., 2001;French validation by Jermann, Van der Linden, d'Acremont, & Zermatten, 2006).The CERQ contains 36 items rated on a 5-point Likert scale (from "almost never" to "almost always") and measures regulation of negative emotions only.The scale assesses the use of nine coping strategies: acceptance, refocus on planning, positive refocus, positive reappraisal, putting the problem into perspective (these five strategies form an "adaptive cognitive regulation" factor), self-blame, others-blame, rumination and catastrophisation (these four strategies form a "maladaptive cognitive regulation" factor).Internal reliability scores for the subscales range from .68 to .87 (Jermann et al., 2006).
Positive and negative state affectivity were assessed through the Positive and Negative Affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988;French translation: Gaudreau, 2000).The PANAS is currently the most widely used measure of affectivity.It consists of 20 adjectives rated along 5point scales, of which 10 assess positive affectivity (PA) and 10 assess negative affectivity (NA).It should be noted that the focus here was on state (current) rather than trait (general) affectivity.Cronbach's coefficients of internal consistency of the positive affect (α = .90)and negative affect (α = .80)are high (Gaudreau, Sanchez, & Blondin, 2006).
Physical health was assessed through the Physical Inventory of Limbic Languidness (PILL; Pennebaker, 1982), which is a list of 54 physical symptoms and bodily sensations.Participants are required to rate items on a 5-point scale (never or nearly never / 3 or 4 times a year / about every month / about every week / more than once a week).The global "somatic complaints" score is highly reliable (α = .90;Mikolajczak, Luminet, & Menil, 2006).
The tendency for excessive and uncontrollable worry was assessed via the Penn State Worry Questionnaire (PSWQ; Meyer, Miller, Metzger, & Borkovec, 1990;French version: Gosselin, Dugas, Ladouceur, & Freeston, 2001).The PSWQ is a commonly used trait measure of anxiety intended to assess a person's overall tendency to experience worry.The PSWQ is a 16-item selfreport scale.The items are scored using a 5-point Likert scale, anchored by "not at all typical" (for me) and "very typical" (for me).The French translation shows very good internal consistency (α = .92)and excellent convergent validity with other measures of worry and anxiety (Gosselin et al., 2001).
Relationship quality was evaluated via a measure created for the purpose of the present study, and inspired by the Network of Relationships Inventory (NRI; Furman & Buhrmester, 1992).Our measure appraises relationship quality through five dimensions: nurturance (5 items, e.g., "to what extent do you help X for things that s/he cannot do by her/himself"), intimacy (4 items, psycho.belg.2011_1.bookPage 60 Tuesday, December 7, 2010 8:34 AM e.g., "to what extent do you share secrets and feelings with X"), conflict (4 items, e.g., "to what extent do you argue with X when you disagree with him/her"), self-respect in the relationship (3 items, e.g., "to what extent do you feel respected in the relationship with X") and respect of others in the relationship (4 items, e.g., "to what extent do you accept that X may have opinions, habits, and projects different from yours").Thus, there were 20 items in total, which had to be answered six times, concerning participants' partner, best friend, friend, most important sibling, father, and mother, respectively.All items were rated on a 5-point Likert scale (1 = little or none to 5 = the most).The alpha of the global relationship quality score was .74 in the present study.
Happiness was assessed using the Subjective Happiness Scale (Lyubomirsky & Lepper, 1999).The measure is composed of 4 items scored on a 7-point Likert-type scale (from 1 = less happy to 7 = more happy).The internal consistency in our sample was good with a Cronbach's alpha (α) of .85.
Job performance was assessed via a measure created for the purpose of the present study and inspired by the Leadership Architect (Lombardo & Eichinger, 2000).This measure is a self-evaluation of managerial competencies.It consists of 14 items rated on 10-point Likert-type scales ranging from 1 (not competent at all) to 10 (fully competent).These items evaluate different managerial skills such as autonomy, motivation, and team spirit.Two additional items assess overall performance ("in general, how competent are you in your work?" and "overall, how competent are you to achieve what is expected from you?").The alpha of the global score was .87.

Factor structure
Three hundred and twenty students completed the ERP-R for factorial structure validation.Participants' scores on each item (scenario) were subjected to principal component analysis.The Scree plot and Kaiser's eigenvalue extrac-psycho.belg.2011_1.bookPage 61 Tuesday, December 7, 2010 8:34 AM tion criteria suggested the presence of three factors (the eigenvalues were respectively: 4.652, 2.040, and 1.039).We applied Parallel Analysis (PA) to our data set, a method which is currently considered the most reliable procedure to determine the correct number of factors (see Zwick & Velicer, 1986 for a comparison of factor retention decision methods, and Hayton, Allen, & Scarpello, 2004 for methodological issues in PA).The eigenvalues and standard deviations generated from completely random data (and necessary to perform PA) were obtained through the "Marley Watkins Monte Carlo PCA for Parallel Analysis" program (Watkins, 2002) using the following parameters: 15 variables (scenarii), 320 participants, 1000 replications.We then compared our observed eigenvalues to the 95 th percentile of the eigenvalues generated from these random data, in order to reject factors that are most certainly artificial (at p < .05).Results suggested to consider only 2 factors.A two-factor solution was found by using the PROMAX algorithm (kappa = 4).The factor pattern matrix is presented in Table 1.An item is judged to belong to a factor if its loading on this specific factor was > .40.If two items were or above .40,we allocated it to the factor with the highest saturation.The two factors were composed according to our theoretical expectations, and were thus labelled accordingly: "down-regulation of negative emotions" and "upregulation of positive emotions".This solution accounted for 42% of the total variance and the strength of the intercorrelations between factors (r = .49)was fully in line with our theoretical assumptions.

Internal consistency
The reliability of the global ERP-R score was good (α = .84).The two factors down regulation of negative emotions and up-regulation of positive emotions also showed satisfactory internal consistencies (Cronbach αs were .83and .79,respectively).

ERP-R and gender
Males scored significantly higher on the factor "down regulation of negative emotions" and on the global ERP-R score than women (see Table 2).***p ≤ 0.001; **p ≤ 0.01; *p ≤ 0.05.

Discriminant validity with respect to cognitive ability (IQ)
Discriminant validity refers to the degree to which scores on a test do not correlate with (are "independent of" or "orthogonal to") variables they are not supposed to correlate with, given the nature of the construct.
Neither the global ERP-R score nor any of its factors correlated with non verbal reasoning and verbal skills (see Table 3).

Convergent validity
Convergent validity refers to the degree to which scores on a test correlate with scores on another test that is believed to measure a closely related construct (i.e., the two tests should end up ranking people in pretty much the same way).

ERP-R and trait emotional intelligence
As shown in Table 3, the ERP-R was significantly positively correlated with trait emotional intelligence (global score and factors).This association seemed mostly attributable to the factor down-regulation of negative emotions.

ERP-R and the Five-Factor Model of personality (FFM)
The ERP-R (global score and the two factors) was positively associated with emotional stability and agreeableness.There was no correlation with extraversion, conscientiousness and openness (Table 3).

ERP-R and the Cognitive Emotion Regulation Questionnaire (CERQ)
The global ERP-R score was positively associated with the use of adaptive cognitive emotion regulation strategies (Table 3).The down-regulation of negative emotions was positively associated with the use of adaptive cognitive strategies and negatively associated with the use of maladaptive ones.

Susceptibility to socially desirable responding
Although the factor "up-regulation of positive emotions" shared no variance with social desirability, the factor "down-regulation of negative emotions" was negatively correlated with social desirability.Yet, the global ERP-R was not significantly affected by socially desirable responding (see Table 3).

Criterion validity
Criterion validity refers to the ability of a test to predict some criterion it should theoretically be able to predict.Criterion-related validity can either be concurrent or predictive.Concurrent validity refers to the correlation between the predictor and criterion scores obtained approximately at the same time.
Predictive validity refers to the degree to which scores on a test predict future behaviour on a criterion variable.

ERP-R and positive and negative state affectivity
As shown in Table 3, the ERP-R is a significant predictor of positive state affect as measured by the PANAS administrated six weeks after the ERP-R.Surprisingly, this effect was driven by the "down-regulation of negative emotions" factor and not the "up-regulation of positive emotions" factor.Neither the global ERP-R score nor any of its factors predicted negative state affect.

ERP-R and mental health
The ERP-R was highly associated with mental health (see Table 3).Eight of the nine dimensions of the BSI were negatively correlated with the ERP-R.
The dimension of the BSI that is the most highly associated with emotion regulation is hostility.At the factor level, regulation of both negative and positive emotions was a significant predictor of mental health.

ERP-R and somatic complaints
Emotion regulation was negatively correlated with somatic complaints (see Table 3).

ERP-R and the tendency to worry
As shown in Table 3, emotion regulation is a significant predictor of the tendency for less worry and less uncontrollable worry.

ERP-R and perceived social support
Emotion regulation showed significant relations with four of the five dimensions of social support, with the highest relationships showing up for conflict and nurturance and somewhat lower but still significant correlations with respect of others in the relationships and intimacy.The factor "down-regulation of negative emotions" was associated with dimensions of conflict, nurturance, and respect of others in the relationships."Up-regulation of positive emotions" was only associated with nurturance and intimacy.

ERP-R and the propensity to experience various discrete emotions
As shown in Table 3, emotion regulation was positively related to the propensity to experience joy and negatively related to the propensity to experience sadness, envy, and shame.This association seemed mostly attributable to the factor "down-emotion regulation of negative emotions", which was positively related to the propensity to experience joy and negatively related to experience sadness, fear, envy, and shame.The "up-regulation of positive emotions" was positively associated with joy and negatively associated with sadness.

ERP-R and happiness
Results in Table 3 show that the ERP-R total score and all subscores significantly correlate with subjective happiness.

ERP-R and job performance
Emotion regulation was positively related to performance (see Table 3).This association seemed entirely attributable to the factor "down-regulation of negative emotions".

Incremental validity
The present study also sought to examine the incremental validity of the ERP-R.Incremental validity is defined as the degree to which a measure explains or predicts a phenomenon of interest, relative to other measures.We chose the tendency for excessive and uncontrollable worry, physical health and mental health as dependent variables, as these variables were also well predicted by emotional stability (first dimension of the FFM) and trait emotional intelligence.Scores on Trait EI, emotional stability, and the ERP-R served as predictors.It was hypothesised that the ERP-R would be a reliable predictor of all three criteria, in the presence of both trait EI and emotional stability.We tested the incremental validity of the ERP-R to predict tendency for excessive and uncontrollable worry over and above Trait EI and emotional stability using a hierarchical procedure (Cohen & Cohen, 1983).To determine whether the ERP-R scores are able to account for additional variance above and beyond scores from the TEIQue and the emotional stability, the TEIQue and the emotional stability were entered in Block 1 and scores from the ERP-R were entered in Block 2. The results are presented in Table 4.The ERP-R significantly predicted the tendency for excessive and uncontrollable worry over and above the TEIQue and the emotional stability.
In order to test whether the ERP-R was able to predict physical health beyond established constructs such as EI and personality, we analysed the incremental validity of the ERP-R over and above the TEIQue and the emotional stability factor of the D5D using a hierarchical procedure.TEIQue and emotional stability were entered as the first block and the ERP-R was entered in the second block.As depicted in physical health over and above the effects of the TEIQue and emotional stability.ERP-R also significantly predicted mental health over and above the TEIQue scores (Table 6).

Discussion
Results lend strong preliminary support to the validity of the ERP-R.First and foremost, internal consistencies of the two factors and the total score were good.Second, factorial analyses yielded a two factors solution, confirming theoretical expectations about the dependence of negative and positive emotion regulation.Moreover, the ERP-R displayed evidence of convergent and discriminant validities with a large number of other variables.Firstly, the two ERP-R factors correlated in meaningful and theoretically congruent ways with trait emotional intelligence, emotion regulation, and the big five factors of personality.Specifically, the ERP-R was significantly positively correlated with trait emotional intelligence (global score and factors).This association seemed mostly attributable to the factor down-regulation of negative emotions.The factor up-regulation of positive emotions was not significantly correlated with two factors of the TEIQue (self-control and sociability).The ERP-R scale regulation of negative emotions showed moderate correlations with the CERQ.This is not surprising because both questionnaires assess the same construct and several regulation strategies are similar in both tests.However, the ERP-R is different from the CERQ, especially regarding the presentation of the items.First, ERP-R is a vignette-based instrument whereas CERQ is a classical self-report questionnaire.Second, ERP-R assesses two forms of emotion regulation whereas CERQ targets only the down-regulation of negative emotions.Finally, eight regulation strategies were presented in the ERP-R, whereas nine were identified in the CERQ.However, the two questionnaires are similar in that they employ common regulation strategies such as acceptance, positive reappraisal, and rumination.Other strategies featuring in the CERQ are not included in the ERP-R, such as catastrophizing and blaming others.The ERP-R (global score and the two factors) was positively associated with emotional stability and agreeableness.There was no correlation with extraversion, conscientiousness, and openness.These results are consistent with previous findings showing that emotional disturbances in subjects are associated with higher neuroticism scores (Izard, Libero, Putman, & Haynes, 1993).They are also consistent with the idea that people characterised by higher agreeableness scores are more likely to express positive emotions because they experience positive relationships (Costa & McCrae, 1992b).Conversely, ERP-R scores were unrelated to non verbal reasoning, which is consistent with the theory that individual differences in typical behaviour in emotional situations are different from cognitive intelligence (Freudenthaler & Neubauer, 2005, 2007).Interestingly, ERP-R scores were found to be influenced by gender.The fact that men scored higher than women on downregulation of negative emotions is consistent with the divergent socialisation psycho.belg.2011_1.bookPage 69 Tuesday, December 7, 2010 8:34 AM of emotion as a function of gender (e.g., "Boys don't cry"; Mikolajczak, 2009) and also with the fact that women are generally more anxious than men (Leach, Christensen, Mackinnon, Windsor, & Butterworth, 2008;Mackinaw-Koons & Vasey, 2000).These gender differences should be taken into account for the establishment of norms and for the interpretation of scores, especially when women and men have to be compared (e.g., organisational or educational settings).
The prediction of "objective" life outcomes such as professional success or health status was beyond the scope of the present study.All criteria examined in the present study were self-reported.These are, however, not trivial criteria, especially given that emotional well-being is a major goal in life for most people.Furthermore, examining such criteria was essential to ensure construct validity.In this respect, the ERP-R predicted a substantial and meaningful part of variance of future positive state affectivity, mental and physical health, perceived quality and quantity of social support, propensity to experience various discrete emotions, self-reported performance, and happiness.These findings are consistent with theory in the emotion regulation field (see Gross, 2007).Although such evidence of predictive validity is a necessary condition for a test to be claimed useful, it is, however, not sufficient.To be deemed useful, the ERP-R should also demonstrate that (1) findings are not attributable to social desirability, and (2) that other tests/questionnaires are not sensible enough, at least not as efficient.With respect to the first condition, responses to the ERP-R are rather independent of social desirability except maybe for the regulation of negative emotions.The two factors are, however, not equally subject to desirable responding.The association between ERP-R and social desirability appears entirely attributable to the factor down-regulation of negative emotions.This might be explained by the fact that some strategies such as substance abuse have a strong negative social connotation.The second condition also seems to be met as the ERP-R predicts a tendency for excessive and uncontrollable worry and physical health over and above trait emotional intelligence and emotional stability, and predicts mental health over and above trait emotional intelligence.The foregoing points represent promising findings in favour of the validity and usefulness of the ERP-R.
Before concluding, several limitations have to be acknowledged.First, most of the participants were students, which restricts range and generalisation, especially regarding "age", for which results have to be interpreted with caution.Future research should be extended to other populations.For example, it would be interesting to validate the ERP-R with an adolescent sample.Emotion regulation is crucial in this period and deficits in this area could lead to psychopathological problems.In fact, this is a period where the higher vulnerability of women towards depression becomes observable.Emotion regu-psycho.belg.2011_1.bookPage 70 Tuesday, December 7, 2010 8:34 AM lation is important in adolescence because the physical and psychosocial transformations experienced during this period are accompanied by strong emotions (d'Acremont & Van der Linden, 2007;Larson & Lampman-Petraitis, 1989) and several neural or cognitive systems that are supposed to control emotion are still maturing (d'Acremont & Van der Linden, 2007;Hooper, Luciana, Conklin, & Yarger, 2004).Also, many forms of psychopathology in adolescence are related to difficulties in emotion regulation.Future studies would also certainly benefit from considering objective criteria, such as behavioural (e.g., pro-social behaviours) or physiological (e.g., cortisol secretion) ones.The ERP-R has the advantage of informing about emotion regulation strategies that people use.However, in this study, we did not consider regulation strategies separately.We focused on the total score and on the two factors.In a future research, it will be interesting to consider regulation strategies separately and to examine the relations between specific strategy and various related variables like mental health and well-being.Also, we chose eight regulation strategies but it has to be acknowledged that other strategies (e.g., coping strategies) exist and have an influence on different outcome variables such as mental health.
Overall, the results show a clear two factor solution for the ERP-R and high correlations with convergent and divergent scales as well as good predictive and incremental validities.The present results provide encouraging evidence in support of the utilisation of the ERP-R.The value of the ERP-R is that it enables clinicians and researchers to assess a wide variety of regulation strategies with a single questionnaire, and the ERP-R evaluates the regulation of both negative and positive emotions.The ERP-R determines which strategies individuals preferentially use for negative and positive emotions.This information is particularly interesting in the clinical field where associations between these strategies and psychopathological manifestations can be explored and linked.Thus, the ERP-R gives a reliable and theoretically grounded assessment of typical emotional regulation ability and may, therefore, be useful for both researchers and practitioners.
Note.Factor Pattern Matrix: Coefficients that should theoretically define each factor are in boldface.

Table 1
Factor Pattern Matrix for the ERP-R items (Promax Principal Axis FactoringFour-Factor Solution)(N = 320)

Table 3
Convergent, discriminant and criterion validity of the ERP-R

Table 5 ,
the ERP-R significantly predicted

Table 4
Hierarchical Regression Analyses testing the Incremental Validity of ERP-R over and above trait EI and emotional stability to predict tendency for excessive and uncontrollable worry (N = 62)
reflète(nt) le plus votre manière de réagir dans ce genre de situation.