TENDENCY TO SELF-HANDICAPPING IN THE SITUATION OF EXPECTED FAILURE

The basic objective of this research is to determine what type of selfhandicapping strategies subjects will use when they face potential failure, under the condition that they have the possibility of strategy choice. Another research objective is directed to the research of relation between the selected self-handicapping strategy and personality characteristics. 120 student of the second and third year of the Department of psychology, University of Novi Sad has been taken into sample. In the first research stage, all students administrared a series of personality self-report tests. In the second experiment stage, the participants were divided into four groups, on the basis of two criteria: the solvability of intelligence test tasks and the presence of hampering factors during task solving. The research results show that there are at least two types of selfhandicapping strategies. One type is used by the majority of people when found in the situation where potential failure is expected. It refers to searching for alibis in external circumstances and it is probably the reflection of sound tendencies of facing possible consequences of information on one’s own incompetence. Another type of strategies presents consistent behaviour pattern, determined by the higher degree of adverse affectivity and negative image of oneself, and it is manifested in the use of poor psychophysical condition as the alibi for failure.

The term self-handicapping was first used by Berglas and Jones (1978), thus explaining one of few known strategies of facing potential failure.When a threat to self-esteem is present, i.e. when failure is expected in some activity, which is most often connected to capabilities, a person actively searches for or creates the factors which might hamper the performance of this activity, and which may serve as justification for potential failure.
The concept of self-handicapping issues from the theories of attributions (Arkin and Baumgardner, 1985;Brown, 1991).Self-serving bias in causal attributions refers to a rather pervasive tendency of individuals to attribute succesful outcomes to themselves and unsecssessful outomes to other factors.By denying personal responsibility for failure, the negative feedback is vastly reduced.Although the failure has occured and cannot be reversed, the implications of the failing outcome in determining one's level of ability is minimized.In contrast, by assuming personal responsibility for successful outcomes, the positive image of one's own abilities can be increased.This distortion in causal attribution is a faurly clear use of attributional priciples aimed at protecting or sustaining one's image.
A more subtle use of attributional principles would consist of performing only certain behaviors or manipulating the context proactively, so that only desired inferences about personal qualities could be drawn.These refined manoeuvers are applied before a given behavioral sequence is undertaken.One of these subtle manoeuvres is self-handicapping.Berglas and Jones argued that this strategy refers to the design of obstacles for the achievement of certain objective and the acceptance of all factors that decrease personal responsibility for mediocracy or failure, and exaggerate personal responsibility for success.In case of mediocre or failling performance, the role of the individual's ability must be discounted as a causal factor, since another plausible cause (the handicap) is present.In the case of successful performance, the role of individual's ability as cause is augmented because the success occured in spite of the adverse circumstances.
Ideal handicap should make impression that it is related with the performance of an activity, but, actually, it should serve as a minimal obstacle to success.For example, tiredness due to sleepless night may serve as the alibi to student for potential failure at his/her exam.If, however, the exam is successfuly passed, despite the handicap, the impression of high abilities is confirmed.
The use of symptoms or personality traits as accounts for future negative evaluation has been extended beyond test anxiety.Some studies have indicated that shyness, depression, hypochondriacal complaints may be used in the same fashion (Arkin & Baumgardner, 1985).Handicaps may be behaviours that decrease the possibility of exact failure cause evaluation.(the decrease of efficiency due to the consummation of alcohol or drugs).Despite these internal handicaps, there are also external handicaps, such as choosing poor peformance conditions or difficult goal choice.
Internal handicaps should carry more negative repercussions for the selfhandicapping individual.For example, to fail because one is drunk should be viewed more negatively than failure attributable to an impossible task.It appears that internal handicap carries a special risk of its own because they ordinarily produce negative personal and social evaluations.However, the disadvantage of external handicaps is that they may be less persuasive than the more costly, internal sort of handicap.They may be more subject to public scrutiny and consequently more easily "discovered" by a skeptical audience.
Two methodological remarks may be made to previous researches of selfhandicapping concept.Namely, for the most part, experimental designs did not provide the possibility of choice strategy could participants would used.
To test whether individuals engage in self-handicapping behaviors when failure anticipated, Berglas and Jones (1978) first led some male subjects to believe that they were likely to succeed on an upcoming test; others were given sufficient reason to believe that future success was unlikely.All subjects were then told that the second part of the experiment involved testing the effect of two new drugs on tests performans.The subjects were then given a choise as to wich drug they wished to ingest.Consistent with tthe notion that individuals will take active efforts to sabotage their performance when future succes is improbable, subjects who believed that success was unlikely were most apt to select the performance-inhibiting drug.In this research, the selection of drug was one possible strategy of self-handicapping.
The practice of the provision of only one possibility has been used in a few replication of the original self-handicapping finding.For instance, Tucker (1981, according to Arkin and Baumgardner, 1985), replicated the Berglas and Jones methodology using alcohol consuption as the dependable measure.
The question that arises after these studies is what type of handicapping strategy an individual is likely to adopt, should he have a choise?
Another objection that may be directed to previous researches of selfhandicapping is selective and insufficient attention paid to personality variables.The concept od self-handicapping was frequently related with self-esteem (Rhodewat, et al., 1991) and self-serving attributional biases (Thompson & Richardson, 2001).At the other hand, anxious persons may use their symptoms as handicap when they are not sure of their success in future tasks (Smith et al., 1982, according to Arkin & Baumgardner, 1985).
Also, self-handicapping may be seen like a personality trait, such as perfectionism (Hobden & Pliner, 1995).
Some people may dedicate high attention to every detail, while preparing for some task, and these activities provided them alibi for failure.Therefore, inefficiency in task performance will be interpreted through their need to do it perfectly.
Here, as well as in the above-mentioned research regarding anxiety (Smith et al., 1982, according to Arkin & Baumgardner, 1985), some personality characteristics and conditions were used as strategies of self-handicapping, while they, also like some other characteristics and conditions are not researched as possible predisposed factors -as the characteristics, which due to their (un)emphasis, make a person more subject to self-handicapping.
The research presented in this paper is outlined with actual objective to answer the questions for which there is serious doubt that they have been neglected in previous self-handicapping studies.The basic objective of this research is to determine which type of self-handicapping strategies will be used by subjects when facing potential failure, under the condition that they have the possibility of strategy selection.
Another research objective is focused to the examination of relation between selected self-handicapping strategy and personality characteristics.Besides selfesteem, as important personality aspect, it was anticipated that some other characteristics may affect the selection of the self-handicapping strategy, which could be permanent and consistent personality characteristic.Also, it was assumed that the presence of more permanent emotional variables, like anxiety and depression features could also be related to the type of selected self-handicapping strategy.

METHOD Subjects
The research was participated by 120 students of the second and third year of psychology, where one class of exercises was replaced with participation in experiment.The average age of participants is 21.
Subjects were divided into four groups, equalised according to gender.Although it was predominantly female sample, each group had proportional number of young women and men (2/1/=2.27;p=0.51).The subjects were also equalised in relation to the intelligence degree, considering the results from entrance examination for psychology studies.Additionally, post hoc control of subject groups equivalence in relation to inteligence was performed.All groups were actually solving a series of tasks of verbal and numerical capabilities, which were the same for all groups.Then, the significances of differences in these control tasks were tested.MANOVA showed that none of possible differences is statistically significant (Table 1 and 2).

Procedure
In the first part of the research, all subjects were pretested with a battery of tests, which controlled relevant and independent variables.The participants were instructed that one of the authors needed the subjects for his research.With regard to the need to coordinate the results of these questionnaire with the entrance examination results, the testing was not anonimous.
In the second part, the subjects solved several intelligence test tasks, after the following instruction: »This is intelligence test designed by your proffessor.You will first do exercisse and several examplesafter that.Ttime for examples is limited to 10 min.Afterwards, you will answer the questions on the back of your paper, and only then shall we return to the rest of the intelligence test.«For the needs of the second stage of the experiment, the subjects were assigned to one of the four experimental conditions, obtained by crossing two criteria: the first criterion was task difficulty (easy and unsolvable), and the second criterion referred to favourable and unfavourable testing conditions.Thus, thirty subjects were assigned to the each of the following groups: with solvable tasks and distractor (SD); with unsolvable tasks and with distractor (UD), with solvable tasks and without distractor (SW), and with unsolvable tasks and without distractor (UW).
The groups that performed the tests under favorable conditions (SW and UW) where tested in the classrooms were one of the research was present.He read the instruction, gave them signal to start, and silently waited for ten minutes, and finely, signalled that the time was up.During all the time of testing no one entered the classroom and the activity was performed in silence.
The groups with unfavourable testing conditions (SD and UD) were invited to participate the research in the hall with TV and video.Having entered the room, the subjects found two researches in it.The researchers were watching cartoons on TV and writing down the aggressive reactions of the cartoon characters into the protocol prepared for them.The subjects were, with apology, informed that both researches must take place simultaneously, because of a problem with premises.The researchers behaved according to the following scenario in the experimental situation: Two researchers are sitting in a hall, doing "their research", not noticing entering students, and without any comments.The third researcher, before reading the test instruction, turns to them.Interviewer: "Dear colleagues, we apologise for this disturbance, but we have some work to do.Would you be so kind to silence TV a bit?"One of the TV wachers: Of course (silences it just a bit).After the instruction for the intelligence test, the communication is continued: Interviewer: "Once again, please silence TV a bit" Researcher: "Well, don't interrupt us all the time!Here, I missed to write something down.I have to rewind it at the beginning" (rewinds at the beginning of the cartoon).Students start to answer the test tasks, and the researchers amplify TV and continue with making notes in the prepared protocol.In Attachment No. 4, there is the protocol filled in during quasi -research.Other elements of testing, explanation, instruction for solving, the time of solving, have been the same as with the groups with favorable conditions.
After they had completed their tasks, the subjects in all four groups were asked to answer few questions.The purpose of questions was to check the manipulation success and self-handicapping measurement.The list of questions is provided in Appendix No. 2.

Instruments
The series of tests for the evaluation of personality aspects, relevant for the research, was used.The series comprised of: The Self-Handicapping Scale (Jones & Rhodewalt, 1982).SHS is 25-items self-report invetory aswered on six-point Likert scale.Cronbach-alpha reliability of our translation of test is only 0.51.This imposed the need to reduce the scale, which is the strategy aslo used by some other researchers (Strube, 1986).So the reliability is slightly improved and amounts to 0.67.Through scale factorisation, three factors describing three different Self-Handicapping strategies were extracted: factor 1: verbal justifications; factor 2: delay, the lack of effort; factor 3: internal selfhandicapping (through psycho-physical state).
The Beck Depression Inventory (Beck et al, 1988) -Level of depressive simptoms was assesed bz BDI, a 21 items self-report inventory.Internal consistency of BDI in our research is Alpha = 0.86.
The Beck Self-Concept Test (Beck et al, 1989).This is a self.reportmeasure some aspects of self-conception.25 traits with 5-point rating scale make up BSCT.The instruction is to rate yourself in relation to other people, for each trate.The original version of the BSCT test, as well as our translation, provide four interpretable sub-scales on the basis of which four fields of self-conception may be evaluated: attractiveness (Alpha = 0.78); working efficiency (Alpha = 0.70); intellectual ability (Alpha = 0.68); and vices and virtues (Alpha = 0.58).Spielberger et al, 1970.)The trait version of Spilberg's anxiety inventory assess anxiety as personality trait.Twenty items of the scale are statements about different forms of anxious reacting, and subjects assess, on four-point scale, how frequently they react in that way.The reliability of our version of the scale is Alpha = 0.84.
"The intelligence tests" consisting of numeric and verbal tasks are designed by one of the authors.They were used as instruments for the experimental manipulation.Tasks examples are provided in the Appendix No. 1.There were two test versions: with easy and with unsolvable tasks.In pilot research, task difficulty was checked.Namely, in the pilot version it appeared that the numerical solvable tasks were too difficult to induce the success feeling.Therefore, final version was significantly facilitated.Verbal tasks were found to be easy enough already in the pilot study.
In both test versions, the series of tasks was the same (and solvable), and they had an additional function of aditional check of the group equalisation according to intelligence.The test solving was limited to ten minutes -the time which, during the pilot research, was found out to be long enough for everyone to answer solvable tasks and enough for the subjects with unsolvable form to be convinced that short time was not the reason of their failure.

The treatment of criteria variables
Through the optimisation of the linear combination of a given questions (pi1/q1 -pi7/q7; Appendix No. 2), latent dimensions with pre-defined structure were obtained.Those latent dimensions served as criteria of self-handicapping in the multivariant analysis of co-variance

The check of experimental manipulation success
Since the basic objective of the manipulation was focused to the inducement of success and failure with the subjects, by providing solvable and unsolvable tasks, variance analysis shows that four groups of subjects significantly differ in relation to the evaluation of their own success (Table 4).Both groups, which had unsolvable tasks, evaluated themselves as less successful than groups that had solvable tasks.Besides that, the groups of subjects with solvable tasks evaluated themselves as successful, but the group with solvable tasks without distracter experienced itself as more successful in relation to the group that solved the tasks in the conditions without distractor.
The results are significant, not only as verification of the experimental manipulation success, but also as indirect significance indicator ascribed to the conditions where testing was performed.Namely, the subjects who were solving easy, i.e. solvable tasks in the conditions with the distracter, evaluated themselves as successful to the greatest extent, with tendency (which still does not reach the level of statistical significance) to experience themselves as more successful even in relation to the subject group who solved the same easy tasks in the conditions without distracter.It is possible that greater value is attached to the success achieved with difficulties, and also the image on one's own capabilities becomes more positive.

The results of the multivariate analysis of covariance
Criteria variables in the analysis of covariance were abandonment, external conditions, and psychophysical conditions.Predictor variables are those that result from the experimental manipulation (task solvability and testing conditions), and personality variants are involved in the analysis as covariances.By testing total model, the results indicating that that the relation of predictor variables with all three types of handicapping criteria is significant at the level p = 0.00 were obtained, which points out the significance of the experimental manipulation and personality variables for the prediction of the selection of abandonment, external and psycho-physical conditions as self-handicapping strategies.Nevertheless, the highest percentage of common variability (46%) with predictors is shown by the external conditions.Probably, the conditions of testing and subject characteristics contributed, to the greatest extent, to the selection of the type of self-handicapping, which was imposed by the experimental manipulation itself, i.e. the task solvability and distracter presence.Contrary to this, the lowest percentage of variability with predictors is shown by the abandonment (31%).By its logic, it probably serves as self-esteem protection, for the abandonment prevents facing not only expected future failure, but also possible success.In the prediction of criteria variables, variables resulting from the experimental manipulation (testing conditions and task difficulty) have been more significant in relation to variables from the perspective of personality (depression and psychophysical condition as the self-handicapping strategy).Predictors that have impact to all three types of handicaps provided in this research are (un)solvability of tasks and psychophysical condition as selfhandicapping strategy.
The selection of the abandonment as the handicap during the experiment is impacted by only one variable resulting from experimental manipulation, which is task (un)solvability, and from the perspective of personality -psycho-physical condition as the self-handicapping strategy.
The selection of external conditions as the handicap in the course of experiment is also impacted by the task (un)solvability and poor experimental conditions.From the perspective of personality, significant variables are psychophysical condition as well as the self-handicapping strategy.
The selection of the psycho-physical conditions as the handicap in the course of the experiment is impacted by only one variable resulting from the experimental manupulation, which is the task (un)solvability.The psycho-physical condition as the self-handicapping strategy and depression occur as the predictors of this handicap from the perspective of personality.The presence of the distractors as a form of the experimental manipulation impacts the fact that external conditions are evaluated as worse.Similarly to this, the task solvability, which has been also varied experimentally, impacts the selection of the abandonment as the self-handicapping strategy, but also that the external and psycho-physical conditions are evaluated as worse.At the same time, this is the most robust factor.The subjects who select the abandonment as the handicap in the course of the experiment, achieve low scores at the attraction dimension of self-concept, i.e. the impression they make on others, and they tend to use the psycho-physical condition as the self-handicapping strategy.
The subjects who select the external conditions as the explanation for failure in the course of the experiment, tend more toward the use the psycho-physical condition as the self-handicapping strategy, to some higher extent.They also demonstrate the tendency to evaluate their working efficiency as high.At the same time, this means that the conditions in which the tasks were solved have been suitable to the subjects who actually evaluate their working efficiency as low.

DISCUSSION
A primary aim of this study was to test the assumption whether subjects when anticipated failure would choose one of the three possible handicaps, consistent with the estimated personality traits.This assumption is confirmed.Namely, both the experimental manipulation and the personality traits emerged as predictors of all three types of handicaps.Possible handicaps during the experiment were the withdrawal, psychophysical state and external circumstances.
The withdrawal presents the final instance on the dimension of lack of effort ,which can serve as a useful impediment to performance and thus obscure ability inferences based upon failing outcomes.The individual who expects to fail will reduce effort and thus minimize the likelihood of an attribution of low ability (Frankel & Snyder, 1978according to Arkin & Baumgardner, 1985).In this experiment, the withdrawal is, as well as other provided handicaps, significantly related to the difficulty of the tasks.The direction of relation points out that those subjects faced with potential failure tend to give up.
Two personality traits predicted withdrawal -internal handicap and low attractiveness as self-concept.This handicap use persons who usually demonstrate the tendency to the self-handicapping, as well as those who evaluate themselves as less attractive in relation to other people.Their estimate that they are less likable may be related to the self-handicapping in two ways.They may select this final measure of handicapping because they do not care what impression they will make, which may be the reflection of the psycho-pathological manner.But, these individuals may, also, manifest their desperate attempt to preserve their own selfesteem preceding unavoidable failure.
In this research, the withdrawal was evaluated through the question referring to further task solving.It is important to emphasise that the subjects who decided not to continue participation in the research did not express this explicitly.Namely, none of them stood up and said that he/she didn't want to continue.After the experimental tasks, all participants continued to wait and expect the rest of the tasks.It is possible that the answer to the question about further participation represents verbally expressed opposition (or aggression).However, this verbal withdrawal is also the self-handicapping strategy, providing the alibi for failing afterwards.By this strategy subjects indirectly demonstrate that they will not make effort during the rest of experiment.
The external circumstances is the self-handicapping strategy, which, according to the previous researches, costs a person the least, but, at the same time, this is also the least convincing strategy (Arkin & Baugardner, 1985).The subject's decision to use the external circumstances as the handicap was influenced by the task solvability and objectively poor circumstances.The external circumstances as the handicap were most frequently used by the subjects who also expressed internal handicaping like trait and high work-efficacy.It is possible that those are the individuals who are self-handicaping only if necessary.These persons believed in their diligence, tidiness, and responsibility.Probably, the research circumstances disturbed them the most.The poor circumstances were the attack to their effort, the need for order and responsible attitude to the task solving.Having high opinion of their work-efficacy, they "rebelled" against the objectively poor conditions, because their virtues could not be expressed.It is possible that this maneuver is used only when it is neccessary, like in this experiment.
The subjects who used the psychophysical state as the handicap, also expressed high score on the internal handicaping like trait.At the same time, they are also more depressive and they achieve high scores at the attractiveness as self-concept.The handicapping criterion regarding the search of the alibi to one's own psychophysical state is not precisly operationalised.Namely, the objective of this research was not to examin individual self-handicapping patterns of the psychophysical state.For this purpose, none of the instruments that would have the function esteemed of the anxious state, tiredness or ilness.This criterion was defined through the subjective subject's experience of current condition.
The subjects with unsolvable tasks, estimated their psychophysical state as significantly worse than the subjects with solvable tasks.This is one of the most significant results of this research.Namely, internal handicaps are far more persuasive than the external ones, but they have higher price, especially when one patern of the handicap is predominated (for example, anxiety).In this research, the individuals who used the psychophysical state as the handicap express high score on the internal handicap like trait, they are more depressive, they have low score on the attractivenes like self-concept, and their self-concept includes some negative traits, i.e. shortcomings.Probably those are the persons with self-handicapping tendency as consistent behaviour pattern in coping with potential failure, and they applied it in this experiment, when they faced unsolvable tasks.
The question remains why do persons with depressive tendencies and with negative self-concept use handicapes when expecting failure?They expressed negative view about themselves, but at the same time they didn't achieve low selfesteem scores.The self-handicapping is primarly a protective strategy, which maintains high self-esteem, so, it is necessary to make distinction between selfconcept and self-esteem.Namely, the self-concept is reffers to the cognitive aspect of self-regard, while the self-esteem implicates both evaluative, i.e. emotional attitude to oneself.It is possible that the subjects who use the psychophysical state in this research are the persons who see their bad characteristics, they are more unhappy due to this, but they also tend to use their misfortune in the aim of selfesteem protection.Thus, their self-esteem is not neccessarly decreased.They may evaluate themselves as unlikable, with a series of negative traits, but, they are still worth respecting.
In this research, the criteria variables are determined by the predictors from the experimental situation, mostly than by the predictors from personality traits.Probably experimental situation have higher role in decision to apply the selfhandicapping strategy than personality traits.However, the psychometrical characteristics of the instruments applied for the evaluation of personality traits are poor.Because present the justifiable doubt that some other instruments of the personality traits would be seen as more adequate in the prediction of this phenomenon.
The participants who use their worse psychophysical state as the alibi for the potential failure, presents reaction to the situation where the failure is realistic, and it is also predisposed by some consistent personality characteristics, such as depression or negative self-concept.It is possible that different forms of the selfhandicapping, as searching for the alibi in the psychophysical state, may present the etiological basis of many psycho-pathological symptoms.Namely, Arkin and Baugardner (1985) conclude that people use attributional principles to manipulate their image by developing symptom, consuming alcohol and drugs, which enable self-serving attributions for the possible failures.
In modern psychopathology theories, the self-handicapping is also seen as the basis of disorders such as hypochondriasis (Smith et al., 1983) and somatomorphic disorders (the presence of complaints where there is no physiological basis).In one research, Baumeister (Baumeister et al., 1990) presented preliminary evidence that some people may use obesity as the self-handicapping strategy.
It is important to make distinction between this type of the handicapping and the strategies people used ad hoc, in the situation when they estimate that they will fail, such as the alibi in this research seen as worse external conditions.These strategies are, to some smaller extent, determined by personality characteristics, i.e. there is no doubt that they are the reflection of pre-formed style of facing the potential failure.Therefore, the problems that occurred during the processing of the results regarding the self-handicapping scale (SHS, Jones & Rhodewalt, 1982), may present the reflection of the inadequate attempt of the conceptualisation of the possible measure of this strategy.It seems that the attempt to operationalise the selfhandicapping concept as the questionnaire, which is more applicable to the evaluation of basic personality characteristics, is not appropriate regarding the problem.Since there is the possibility for the situation characteristics to determine this strategy, some form of S-R scale could perhaps present more appropriate form of the instrument.

CONCLUSIONS
When anticipated the failure, people used different forms of the selfhandicapping strategies.The choose depends on the situation characteristics, but also predisposed personality traits.
It may be concluded that there are at least two types of the self-handicapping strategies.One type usually used in the situation where the people expected the failure.This is the search for the alibi in the external circumstances (if there is possibility for this) and it is probably the reflection of sound tendencies of facing possible consequences of information on their own incompetence.The second type of the strategies presents the consistent behaviour pattern, determined by the higher degree of negative affectivity and negative image of oneself, and it is manifested in using bad psychophysical state as the alibi for the failure.The tasks in some future empirical researches of this concept should be focused to the more adequate operationalisation of the psychophysical state like strategy of self-handicapping.