Be Careful What You Wish For ! Desired and Actual Behavior Inconsistency in Frustrating and Provoking Situations as Predictors of Depression

Studies have shown that both expressing uncontrolled anger and suppressing anger are related to depression. Taking the latter into consideration, we explored another possible mechanism that could be used to predict the level of depression – the consistency between desired and undertaken behavior. In Study 1 (N = 270) we tested whether depressive symptoms are related to emotional reactions in provoking or frustrating situations. The results revealed that higher levels of sensitivity to provocations and to frustrations contributed to higher levels of depression. In Study 2 (N = 195) we applied quantitative and qualitative analyses to test whether the severity of depressive symptoms could be predicted by the consistency of what people desire to do and how they actually act, when exposed to frustrating and provocative situations. The results showed that higher perceived consistency was related to lower levels of depression. The results are discussed in terms of the psychodynamic theory.


Introduction
Depression is one of the most widespread mental disorders -the World Health Organization (2016) estimated that globally 350 million people suffer from this illness.There are several theories regarding the psychological, biological, and social factors underlying depression.Some theories, such as the psychodynamic approach, postulate that depression is associated with a loss that a person experienced in the past, but did not confront.It is possible, thus, that the lower mood and sadness cover angry feelings that were never expressed (McWilliams, 2011).Similarly, it has also been suggested that depression, at least in women, is related to the way anger is experienced and processed (Droppleman & Wilt, 1993;Sperberg & Stabb, 1998).It is suggested that women's higher susceptibility to depression, as compared to men, could be explained by women's stronger tendency to exhibit hostile behaviors or suppress angry feelings, instead of expressing their anger straightforwardly (Biaggio & Godwin, 1987;Goldman & Haaga, 1995).Some researchers, however, claim that women do not have a stronger tendency to suppress anger than men do (Biaggio & Godwin, 1987;Kopper, 1993).This inconsistency in research results could be related to cultural disparity in what men and women are allowed to do in certain social contexts (Markus & Kitayama, 1991;Thomas & Atakan, 1993).
Manifestation of angry feelings depends not only on gender (Archer, 2004), but also on the perception of a particular situation (Ajzen, 1991;Markus & Kitayama, 1991).Ajzen (1991) stressed that human behavior should be analyzed in a specific context.Attitudes towards planned behavior, perceived behavioral control, and subjective norms influence our intentions and the actual behavior.If there are any problems with control, consistency of intention and behavior could be diminished (Ajzen, 1991).However, intention is not equivalent to desire (Malle & Knobe, 2001).Desire does not determine whether someone will take action to make it become reality.It has been revealed, for example, that symptoms of depression are a consequence of cognitive distortions in how people see and understand causes of particular events and situations (Golin, Sweeney, & Shaeffer, 1981).
Researchers also stress that there is a great differentiation among the various types of experiences associated with depression, and that it is related, among many factors, to individual differences.Some results indicate that dependency and self-criticism are primary dimensions with a potential to differentiate how people experience depression symptoms (Blatt, Quinlan, Chevron, McDonald, & Zuroff, 1982).Such results were consistent in both clinical and nonclinical samples (Blatt et al., 1982).Those factors are related to both behavioral and psychoanalytic understanding of depression formulations (e.g., Beck, 1967;Freud, 1917Freud, /1957;;Seligman, 1975).On one hand, there is helplessness or dependency, and on the other, negative feelings about the self and reality, and those two are central issues in depression.It is also possible that both factors can coexist, which is consistent with the theory of Melanie Klein (2002), stressing the dynamic aspect of human experience and development.She described two development positions of human identity, namely the paranoid-schizoid and depressive positions.Depending on personal and contextual factors, once they have been successfully developed they can be activated.In case of the paranoid-schizoid position, which is developed earlier in ontogenesis, the superego (that is, the Freudian "internal god") is very harsh and is punishing the individual for what are perceived as destructive aggressive impulses (Klein, 2002).In this position, children are very dependent on their parents, and have problems tolerating any frustration, because the self and the feeling that they are a distinct human individual have not developed yet (McWilliams, 2011).In order to be able to develop the depressive position, a secure attachment needs to be created with the attachment figure (usually a parent or a main caregiver).The depressive position is related to the ability to experience guilt and to think and reflect, because in case of the paranoid-schizoid position, morality is very archaic, with an eye-for-an eye prerogative.This analytic theory is consistent with more current models, such as psychosocial development first described by Erickson and colleagues (e.g., Erickson et al., 1985;Rosenthal, Gurney, & Moore, 1981).Dependency seems to be strongly related to the earlier stage of human development, and the depressive position resonates more with guilt and criticism (Agazarian & Gantt, 2003).Moreover, being able to deal with guilt and criticism is positively related to self-clarity, because the greater the self-clarity, the lower the sensitivity to frustration (Lawrence, 2006).
Furthermore, the classical models of depression propose that those who make stable, global, and internal attributions of failures are more prone to depression (Abramson, Seligman, & Teasdale, 1978).Depressed patients also experience more intense hostility, especially inwardly directed, that can also influence the way they perceive social situations and block them from expressing angry feelings openly (Biaggio & Godwin, 1987).The hostility is possibly intensified by the fact that depressed individuals are more sensitive to negative events, possibly due to higher levels of neuroticism (Hettema et al., 2006).Neuroticism also relates to higher individual sensitivity to provoking and frustrating situations (Kendler, Kessler, Neale, Heath, & Eaves, 1993;Zajenkowska, Jankowski, Lawrence, & Zajenkowski, 2013) as well as to increased experience of anger and hostility (Sharpe & Desai, 2001).Moreover, interpreting other people's behavior as more provocative than it is, for example by attributing "victims" of physical assault -who had in reality ostensibly provoked their attacker -as being more provocative, is stronger in people who have higher sensitivity to provocation (Lawrence & Hodgkins, 2009;Lawrence & Hutchinson, 2013) or higher hostility (Sanz, Garcia-Vera, & Magan, 2010).It has also been shown that when people with high sensitivity to provocation feel that they are being deliberately provoked, they tend to act aggressively (Lawrence & Hutchinson, 2013).In an experimental study, people high in sensitivity to provocation delivered louder noise blasts to a bogus partner only when provoked (Lawrence & Hodgkins, 2009).
McWilliams (2011) nonetheless suggests that it is typical for depressed individuals not to experience spontaneous anger and, if anger is experienced, to have a guilty conscience.This is in line with empirical studies showing that in women, both anger-in and anger-out were related to higher depression scores (Thomas & Atakan, 1993).Therefore, it is possible that the cause of depression is not the way people process anger (in or out), but how they perceive particular social situations (e.g., as frustrating or provoking), how they feel in those situations (e.g., sad or angry), and how they act in the face of this or that particular situation.
Most of the conclusions presented above were, however, based on the results from questionnaire studies, and it is often suggested that results which come from this kind of measure should not be taken for granted (e.g., Schmitt, Realo, Voracek, & Allik, 2008).That is why it is vital to capture both qualitatively and quantitatively how depressive symptoms relate to the way individuals associate their thoughts or actions with particular situations, and whether they experience consistency between their thoughts or desires and their actions.
Based on the previous results, suggesting that depressed patients are more hostile and have problems dealing with angry feelings (Biaggio & Godwin, 1987), we assumed that higher depression would be related to higher sensitivity to frustrating and provoking situations.To test this, we referred to Lawrence's (2006) concept of sensitivity to provocation (SP) and sensitivity to frustration (SF).The former relates to feeling aggressive in reaction to goading from others, and the latter to feeling aggressive in response to having one's goals blocked and in response to uncontrollable negative events.Both provocations and frustrations are considered triggers of aggressive behaviors (e.g., Anderson & Bushman, 2002;Berkowitz, 1993;Bushman & Baumeister, 1998).In her approach, Lawrence (2006) wanted to show that although both types of triggers elicit aggressive reactions, there are individual differences in response to these situational triggers (some might be more sensitive to provocations and others to frustration).Although SP and SF are intercorrelated, studies show that they are separate constructs, and the factorial structure of the Situational Triggers of Aggressive Responses scale measuring those constructs has been confirmed in five countries (Mylonas, Lawrence, Zajenkowska, & Russa, 2017).SP and SF have different personality correlations, with SF correlating with lower self-concept clarity, and SP with higher levels of narcissism (Lawrence, 2006).In non-clinical samples, higher SF and SP were related to higher Neuroticism, lower Agreeableness, and -only SP -to lower Openness to Experience (Zajenkowska et al., 2013).SP correlates positively with physical and verbal aggression and with the tendency to act aggressively towards those who provoke (Lawrence, 2006;Lawrence & Hutchinson, 2013).Some studies show that women are more sensitive to provocations than men (Zajenkowska, Mylonas, Lawrence, Konopka, & Rajchert, 2014).However, at the same time, women are not only perceived as less aggressive but also actually behave less aggressively (Archer, 2004).Culture is "blamed" for not letting women express anger directly, but rather encouraging the trivializing or pathologizing of their emotions.At the same time, women are also more prone to depression (Sperberg & Stabb, 1998;WHO, 2016).Thus, it is possible that the lack of congruency between emotions felt and displayed is an important factor leading to depression.Therefore, we assumed that depression could be related to the discrepancy between behavioral desire and actual behavior.In contrast to SP, SF is not linked to externalized aggression (Lawrence, 2006).It was revealed that SF was associated with a tense/tired state (Zajenkowska, Zajenkowski, & Jankowski, 2015), which is described as a mix of fatigue, nervousness, and anxiety (Thayer, Friedman, & Borkovec, 1996), and as such is often accompanied by sadness and negative thoughts about oneself.Thus, we also aimed to test whether situations that are normally considered as triggers of aggressive feelings would also awaken sadness in depressive participants.If supported, this would be in accordance with the assumption that depressive patients may manifest sadness and, at the same time, feel anger which they are not able to express (McWilliams, 2011).That would also be related to the conflict often experienced by patients in a psychotherapy setting, whose desires also make them feel guilty, thus they proceed with defensive behaviors (McWilliams, 2011).
Summing up, two studies were designed to test whether depression is related to sensitivity (both awaking anger and sadness) to provo-cation and frustration, as well as to test whether depressive participants experience inconsistency between desired and undertaken behavior.

Participants
The sample was non-clinical and consisted of 270 young adults, 198 women and 72 men, aged 18 -40 (M = 22.89; SD = 2.06).Data were collected online and the participants were recruited via social media.Most, but not all, participants in the final sample were students and were living in major urban areas.The history of depression or psychiatric treatment were not measured.

Materials & Procedure
All participants were informed of the nature, purpose, and anonymity of the study, and were asked to complete the Situational Triggers of Aggressive Responses (STAR) scale (Lawrence, 2006).The STAR scale was used to measure aggression-related sensitivities, and comprises two scales: sensitivity to frustration (SF: 10 items, e.g., "I feel aggressive when someone ignores me") and sensitivity to provocations (SP: 12 items, e.g., "I feel aggressive when someone makes offensive remarks to me").Participants responded to items on a five-point scale (1 = not at all true for me; 5 = very true for me).Both scales previously showed good internal reliability (SP α = .78;SF α = .77)(Zajenkowska et al., 2013).The internal consistency of the STAR dimensions in the present research was high (SP α = .86;SF α = .84).
Depression symptoms severity was measured using the Patient Health Questionnaire (PHQ-9) (Kroenke & Spitzer, 2002), which is a self-report instrument consisting of the criteria on which the diagnosis of DSM-IV depressive disorders is based.It has a dual purpose: to establish provisional depressive disorder diagnoses, and to grade depressive symptoms severity.In the current study, it was used as a measure of depression severity with possible scores ranging from 0 to 27, where scores ranging from 0 to 4 indicate lack of depression, 5-9 mild depression, 10-14 moderate depression, 15-19 moderately severe depression, and 20-27 severe depression.Its validity has been examined and supported previously (Kroenke & Spitzer, 2002).The internal consistency of PHQ-9 in the present research was high (α = .85).

Results & Discussion
In order to test if SP and SF predict levels of depression, a series of regression analyses was conducted.It was decided to follow the simple regression procedure to avoid the problem of multicollinearity between SF and SP.According to widely used criteria, independent variables with a bivariate correlation exceeding .70 should not be included in multiple regression analysis (Tabachnick & Fidell, 2001), as happened in this case, r (267) = .79,p < .01.The role of gender was analyzed as well.As shown in Table 1, both SP and SF significantly predicted levels of depression.Higher levels of SP and SF contributed to higher levels of depression.
However, the effect was stronger for SF.It was also found that women had higher levels of depression.There was not a significant interaction between either gender and SF (B = -.002,SE = .004,β = -.101,p = .635)or gender and SP (B = -.004,SE = .004,β = -.303,p = .249),suggesting that the effect of SP and SF on depression levels did not depend on gender.
Obtained results are consistent with some other studies showing that trait anger differentiates between subjects high and low on depression scores, so that those with a high level of depression declared a higher trait anger level (Biaggio & Godwin, 1987).People high in trait anger are likely to manifest more frequent and intense displays of anger than those, low in this trait, and they are also more sensitive, especially to frustrating situations (Lawrence, 2006).Our study also confirms a direct association between sensitivity to frustrating situations and depression.Some previous findings also showed that depressed patients present more hostile behavior than comparable healthy subjects, and that could be related to the fact that they perceive social situations as more provoking (Zajenkowska & Konopka, 2015).The current study also confirms a direct association between the sensitivity to provoking situations and depression (Lemaire & Clopton, 1981).
Depressive patients present problems with the expression of anger and often manifest sadness instead, due to internal psychological con- .03Note.Log 10 transformation of dependent variable was used, as depression scores were positively skewed, therefore logarithmic transformation was appropriate (e.g., Goilnt et al., 1981;Benoit, 2011), significance levels: *p < .05;**p < .01;***p < .001flicts.These conflicts are related to low clarity of self, which is also linked with sensitivity to frustrating situations (Gabbard, 2014;Lawrence, 2006).Because of that, we wanted to explore whether frustrating and provoking situations would elicit not only anger, but also sadness, and whether this would predict a higher level of depression.Furthermore, we aimed to test whether depression would be related to the discrepancy between behavioral desire and actual behavior.

Participants
The sample was non-clinical and consisted of 195 participants, 147 women and 48 men, aged 18 -40 (M = 22.90; SD = 3.16).Data were collected online, and the participants were recruited via social media, with the help of a female research assistant.The final sample consists mostly of university students from an urban area, and of young adults who recently graduated.The history of depression or any psychiatric treatment was not measured.This convenience sample is generally younger, more educated, and more urban than the general sample.

Materials & Procedure
In common with Study 1, all participants were informed of the nature, purpose, and anonymity of the study, and were asked to complete the STAR scale.However, we asked participants not only to rate how angry each situation (e.g., when "a friend betrays me") made them feel, but also how sad they typically felt in the described situations.We also asked them to recall such situations from their own experience and to briefly describe: 1) how they reacted in that situation (an open-ended question), 2) whether their reaction was consistent with what they desired to do in that situation (yes/no), and 3) if it was not consistent, what they desired to do (an open-ended question).Participants were also asked to complete the PHQ-9.Again, both instruments were characterized by high levels of internal consistency (Table 2).

Quantitative Analysis Results
To cross-validate whether provocation and frustration, tested by more specific measures, contributed to higher levels of depression, a series of individual regression analyses was conducted to avoid a problem of multicollinear- ity between the examined constructs.The correlation coefficients for the four constructs ranged from between r = .48and r = .78.As is listed in Table 3, higher levels of feeling angry in reaction to provocations (provocation anger) and to frustrations (frustration anger), but also of feeling sad in reaction to provocations (provocation sadness) and to frustrations (frustration sadness), were related to higher levels of depression.The role of gender was also tested; however, this relationship was not significant.
The role of perceived consistency between desired and undertaken behavior was tested by a series of simple regression analyses.The examined constructs were quite highly correlated, r(182) = .64,p < .01.It turned out that both con-sistency in provocative situations and consistency in frustrating situations successfully predicted levels of depression.Higher perceived consistency was related to lower levels of depression (Table 4).

Qualitative Analysis Results
The main goal of the qualitative analysis was to establish whether depression was related to the lack of consistency between participants' actual reactions and their desired reactions in the face of provocative and frustrating situations.To capture the core of this relationship, the answers of 35 participants with severe and moderately severe depression (PHQ-9 score of 15 and above) were compared to the answers of .01Note.Log 10 transformation of dependent variable was used, as depression scores were positively skewed, therefore logarithmic transformation was appropriate (e.g., Goilnt et al., 1981;Benoit, 2011), significance levels: *p < .05;**p < .01;***p < .00135 non-depressed participants (PHQ-9 score below 5).The answers on four STAR questions were analyzed: two questions concerning provocative situations (i.e., "A friend betrays me", and "I am the subject of a practical joke") and two concerning frustrating situations (i.e., "I have academic or work problems" and "I experience family dispute").These questions were chosen because the difference in consistent and inconsistent reactions of depressed and non-depressed participants was the biggest for them.Furthermore, in both types of situations (SP and SF) one of the questions addressed important actors, such as family members or friends.

Table 4 Individual multiple regression coefficients. The log-linear models of depression regressed on consistency of emotional response to provocation and frustration
"A friend betrays me" question.As listed in Table 5, most of the participants from both groups admitted that they had experienced a friend's betrayal.A marked difference is visible, however, in the rate of healthy and depressed individuals who declared inconsistency between their actual behavior and their desired actions.Furthermore, a different pattern of reactions in these two groups was revealed.Although most of the non-depressed participants who declared consistency (72.22%) did not avoid confrontation and had a sense of agency, a minority withdrew from the situation (e.g., did nothing, went to sleep, drank alcohol) or behaved highly emotionally (e.g., cried).
In contrast, almost all depressed participants who declared inconsistency of behavior and desire (93.3%) avoided confrontation and suppressed their emotional reactions instead of "showing anger" or "telling him/her what I really think".These participants also described their aggressive fantasies towards a friend (e.g., "I wanted to bite off her femoral artery") or described a strong sense of guilt, but instead they did nothing, or cut off contact with that person.Half of the depressed participants, who said that they did as they wished to do, described nonconfrontational, evasive behaviors (e.g., drinking), or behaviors aimed at keeping this relationship (explaining his/her emotions, proposing compromise, etc.).The other half displayed proactive actions aimed at resolving the situation.
"I am the subject of a practical joke" question.Almost an equal number of non-depressed and depressed participants were the subject of a practical joke.It was revealed again that most of the non-depressed participants did not find their behaviors and desires inconsistent.Their behavior seemed resourceful and socially adequate ("I know my worth so I laughed at myself with other people"; "I told them what I really thought about this joke").Non-depressed participants who declared incongruence usually decided to avoid any action and suppress the anger felt.
Almost half of the depressed participants acted differently than they desired.Most of them wished to act assertively or even aggressively and to retaliate, but they withdrew from the confrontation.Half of the depressed participants who declared cohesion of desired actions and actual behavior presented proactive reactions.However, they tended to describe their activity rather as attempts, whilst the nondepressed participants described their activity as accomplished acts.The other half of depressed, consistent participants withdrew from the situation (e.g., remained silent) or acted emotionally (e.g., cried)."I have academic or work problems" question.Academic or work problems are equally common in both depressed and non-depressed participants.Again, almost all non-depressed participants presented congruence between desires and actual behaviors.These were, again, mostly proactive behaviors aimed at problem solving.Most of the non-depressed participants with inconsistent behavior thought about giving up at first, but eventually acted constructively (e.g., finished their dissertation).Only one person from this group withdrew from any activity.
As for depressed participants, almost half of them presented behaviors inconsistent with their desires.The majority of the latter group (58.33%), similarly to non-depressed participants, wanted to give up at first, but eventually took action; 33.3% withdrew from any action, and 8.3% behaved emotionally whilst they actually wanted to take action.
"I experience family dispute" question.An equal number of depressed and non-depressed participants experienced family disputes.However, more non-depressed than depressed individuals declared actual-desired action congruence.Non-depressed and depressed participants took similar types of actions, but with different frequency: they tried to mitigate the dispute (non-depressed: 63.63%; depressed: 11.11%), took an active part in the dispute (nondepressed: 27.27%; depressed: 11.11%), or withdrew from it (non-depressed: 9.1%; depressed: 77.78%).Inconsistent healthy participants re-vealed one of two patterns: they either wanted to act calmly but behaved emotionally, or conversely, they wanted to show their anger but suppressed it.Inconsistent depressed participants usually wished to act constructively but they acted emotionally, crying or showing anger (75%).Almost 19% of them retreated from the dispute, even though they wished to either take part in it or resolve the situation.Finally, 6.25% of inconsistent depressed participants wanted to retreat or get their way, but instead they provided more arguments.
Summing up the results of the qualitative analysis, it can be clearly seen that non-depressed participants showed stronger consistency of desires and actual actions compared to depressed individuals.The former group also declared more constructive and proactive behaviors, which were aimed at solving or mitigating the problem.Furthermore, their declarations also suggest that the inconsistent behavior was often of their own choice.

General Discussion
The present research applied quantitative as well as qualitative analysis to explore factors related to depression.In Study 1, we revealed that higher depression was associated with higher sensitivity to frustration and to provocation.In Study 2, by referring to previous empirical studies and to the psychodynamic theoretical approach, we confirmed the assumption that not only anger, but also sadness elicited by frustrations and provocations, were predictors of depression.The current results also showed that the discrepancy between behavioral desires and actual behavior in situations eliciting anger and sadness could be a factor predicting higher depression.To better understand the inconsistency of desired behavior and actual behavior, we added qualitative analysis of the respondents' "flood of associations" (Gabbard, 2014).
Environmental factors seem to be more important than genetic factors in depression development (Gabbard, 2014).Recent stressful events are the strongest predictors of depressive symptoms (Kendler et al., 1993).It is also suggested that the type of event may play a role in modifying the depressive symptoms (Blonski, Conradi, Oldehinkel, Bos, & de Jonge, 2016).Our results stress the importance of acknowledging individuals' perceptions of the type of situation, and indicate that individuals' sensitivity to provocations and frustrations coexist with higher depressive symptoms.It is possible that depressed individuals perceive the same event as more stressful in nature than nondepressed participants.This would be in line with other findings suggesting distortion of the attribution process in depressed individuals (Huang, Hwang, & Ko, 1983;Robins, 1988).Higher hostility, typical for depressed patients, is also related to a more hostile attributional style (Biaggio & Godwin, 1987).People with a high level of hostility have negative and destructive attitudes towards others and a high expectation that others are going to intentionally cause them harm (Sanz et al., 2010).That could serve as an explanation of the higher sensitivity to provocations among depressed individuals.
On the other hand, people who are more sensitive to frustrating situations lack self-concept clarity (Lawrence, 2006), which is related to internal conflict between what a person desires and what he or she allows himself or herself to do (McWilliams, 2011).This is also consistent with our results, where SF was a stronger predictor of depression than SP.
Apart from anger, sadness is a normal and healthy response to any misfortune.However, if this emotion is strong and is felt often, it could indicate depression (Bowlby, 1980).The present results showed that frustrating and provoking situations could not only evoke aggressive thoughts and desires, but could also be con-sidered as a kind of adversity awakening sadness.
Both quantitative and qualitative analysis of how participants desired to react and how they actually reacted, when facing frustrating or provoking situations, revealed that depressed participants more often described feelings of strong negative, mostly aggressive emotions which, however, were not translated into any action.They seemed to concentrate on their own inner states rather than on problem solving.This is consistent with previous studies revealing that depressed individuals have a diminished sense of control over their anger (Biaggio & Godwin, 1987), and this emotion is suppressed or expressed in ineffective ways (e.g., by crying) (Sperberg & Stabb, 1998).This is also in line with Bibring's (1953) description of the mechanism of depression, suggesting that depression is the emotional expression of a state of egohelplessness and ego-powerlessness.Depression occurs when people have some narcissistic aspirations or goals that cannot be fulfilled.In this sense, depression is an ego reaction to narcissistic frustration.That is why maintaining a balance between inwardly and outwardly directed hostility could protect against the development of severe depression (Lemaire & Clopton, 1981).

Limitations and Future Directions
Although it was revealed that the severity of depressive symptoms could be predicted by the consistency of what people desire to do and how they actually act when exposed to frustrating and provocative situations, more research is needed to confirm this result.First, the descriptions of situational contexts were chosen based on their importance as well as universal character (e.g., commonality across cultures).However, we are aware that these results could be interpreted only within the chosen four situational contexts.Although this does not exclude the possibility that the obtained results describe universal mechanisms, future studies could test this issue by applying an experimental design and/or a less artificial setting.Furthermore, in Study 1, but not in Study 2, gender was a significant predictor of depressive symptoms, which is consistent with previous studies revealing women's higher susceptibility to depression (e.g., Sperberg & Stabb, 1998).However, because our study sample consisted mainly of females, gender comparisons should be considered with caution and addressed in future studies.
The current study tried to combine different approaches: psychoanalytic, empirical, and social, which all assume that dependency is related to depression.Our results could also be interpreted in this way because the aggressive impulses are perceived as destructive and therefore cannot be expressed, possibly due to an overly harsh superego.In the group analysis approach, the first phase of the development of a therapeutic group is related to dependency.Patients in a group, regress to a paranoid-schizoid position and the need for dependency is activated.Later, the need for dependency can be analyzed, which can enable patients to develop a well-structured ego-syntonic (observant) self.It would be valuable in future studies to check if depressive symptoms decrease after such an intervention.Furthermore, future studies could test whether such a therapeutic approach is more effective in an "analytic" type of depression (Blatt, 1974) related to the fear of being abandoned and wishes to be cared for, loved, and protected.
The present studies aimed at adding to the understanding of the mechanisms involved in causing depression by focusing on desired and actual reactions to situations that could be perceived as provoking or frustrating.The results suggest that the intensification of unexpressed sadness and anger in provocative and frustrating situations could increase depressive symp-toms.This may be understood as an illustration of internal conflict, where desires from the id level meet norms from the superego level.Such a conflict, coupled with a lower level of control over our feelings, could lead to depression (Biaggio & Godwin, 1987).

Table 1
The Log -linear models of depression regressed on provocation, frustration, and gender -individual multiple regression coefficients

Table 2
Internal consistency, means and standard deviations of STAR scales and PHQ-9 scale

Table 5
Desired action-actual behavior inconsistency rate in healthy and depressed participants