The Impact of Locus of Control on Moral Behavior
Master thesis, MscHRM
University of Groningen, Faculty of Economics and Business
2
Abstract and key words
3
TABLE OF CONTENTS
I. INTRODUCTION ... 4
1.1 Moral and Immoral Behavior ... 5
1.2 Moral Behavior, Immoral Behavior, and LOC ... 6
1.3 Gender and Morality ... 9
II. METHODS ... 11
2.1 Participants ... 11
2.2 Design and Procedure ... 11
2.3 Measures ... 11 2.3.1 Locus of control. ... 11 2.3.2 Immoral behavior. ... 12 2.3.3 Moral behavior. ... 12 III. RESULTS ... 13 3.1 Descriptive Statistics ... 13 3.2 Hypotheses Testing ... 13
IV. GENERAL DISCUSSION ... 15
4
I. INTRODUCTION
The discussions about morality, and in particular about moral behavior, have existed since the times of Aristotle and Plato. Currently, this theme is highlighted in the context of organizational behavior, which in turns constitutes one of the performance predictors and leads to achieving profit of the organization (Campbell, 1990). Rubin, Dierdorff and Brown (2010) highlight that ethic also plays a significant role in a career development, more precisely in a promotion possibilities.
Economic systems deal with immoral, unethical and dishonest behavior. The desire of the firms to achieve their profit by cheating can lead to a destruction of the firm in comparison with those firms, which engage in more ethical behavior (Cooke, 2007). The immoral behavior can lead to lower productivity (Varca & James-Valutis, 1993). The vital importance of cheating for business gives a possibility to observe the problem of finding a balance between ways of achieving individuals’ desires with the ethical and moral aims of organizations. Thus, in this context it is logical to raise the question about the possible individual difference predictors of moral and immoral behavior. In the current investigation, I examine Locus of Control (LOC) as one such predictor.
LOC (Rotter, 1966) is an individual’s beliefs in his possibilities to control his life. Thus, when a person attributes his actions, faults, or successes to his own agency, he is considered to have an Internal LOC. In contrast, an external LOC is characterized by the belief that actions are due to circumstances outside of an individual’s control (e.g., destiny, chance, or luck).
5 relationship between LOC and personal perception of reality. Thus, comparing a person with an external LOC and a person with an internal LOC, the former is more likely to explain his outcomes based on his own behavior. From this point forward in the thesis, I will refer to those with an internal LOC as internals and those with an external LOC as externals.
This current study examines the relationship between LOC and another type of behavior related to personal responsibility: different types of moral (charitable donations) and immoral (cheating) behavior. In addition, there is evidence of a relationship between gender and morality. Taking into account this ambiguity, I will also examine if there are any differences in the moral/immoral behavior based on gender.
1.1 Moral and Immoral Behavior
Moral behavior is defined as acts for production of fair outcomes (Schulman, 2002). School (2008) emphasizes the production of positive outcomes for others, equating moral behavior to prosocial behavior or to a source of motivation for identification goals. Thus, the main feature of moral behavior is altruism.
6 As other scholars have done in the past (Jordan et al., 2011; Jones, 1991), I use the terms moral/ethical and immoral/unethical interchangeably.
1.2 Moral Behavior, Immoral Behavior, and LOC
Before exploring the possible relationships between moral behavior, immoral behavior, and LOC, I want to focus on LOC. I also want to briefly discuss the origins of LOC. LOC is developed from an individual’s early family environment. Thus, the precondition for the development of internals is warm, caring, consistent and sensitive environment. In contrast, unpredictable, inconsistent, and a relatively unfavorable environment creates a feeling of personal helplessness, forming an external LOC (Yalom, 1980). Interestingly, birth order also can play a significant role in the formation of LOC; firstborns more frequently possess internal LOC, because they are more frequently responsible for the household affairs and have to look after younger children, forming sense of self-responsibility (Yalom, 1980). LOC becomes a stable personal quality within the process of socialization.
7 behavior, people expect the same outcomes of their own behavior. Accordingly, individuals with internal LOC were more likely to link their behavior to specific outcomes than those with external LOC. Trevino and Youngblood (1990) found that LOC was related to ethical behavior through its influence on outcome expectancies. Specifically, they found that individuals with internal LOC are more likely than individuals with external LOC to behave more ethically, because they recognize the rewarded behavior of others, and adopt their own behavior to desirable outcomes. It happens due to ability of internals to understand the coherent relationship between their behavior and outcomes. This meant that outcome expectancies, such as possible rewards or punishment of the behavior, likely play an appropriate role in the relationship between LOC and moral behavior.
The authors also took into account the conception of cognitive moral development (CMD; Kohlberg, 1969). Trevino and Youngblood (1990) found that CMD moderated the relationship between LOC and moral behavior, meaning that individuals at the principled stages of CMD with internal LOC behaved more ethically then individuals at lower stages of CMD.
8 principles of morality (Lefrançois, 2000). On this level a person makes his first attempt to clarify the sense of moral rules, he tries to make his own framework of morality. Thereby, individuals obey the general social moral norms, according to their own norms (Lefrançois, 2000).
Leftcourt (1976) also linked moral behavior with LOC. Thus, he stated that the likelihood of resistance to social pressure, as well as resistance to temptation, was higher in internals’ behavior than in externals’ because externals are more likely to be sensitive and responsive to the influencer’s status and, thus, they are more ready to accept the directions of the influencer without manifestation of the resistance. Internals see the possibility to accept the direction or suggestion of the influencer only in case when there is a congruence of this suggestions and directions with their own view and self-directions. Thus, individuals with internal LOCs have higher level of resistance. Midlarski (1971) found support for the relationship between LOC, helping behavior, and altruism. Specifically, internals were more likely to demonstrate altruism and helping behavior than were externals.
9 both cognitive traits (internal LOC and field-independence) leads to high levels of individual morality.
In the present study I attempt to examine the relationship between LOC and moral behavior. Thus, I test following hypotheses:
Hypothesis 1: External locus of control is positively correlated with immoral behavior. Hypothesis 2: External locus of control is negatively correlated with moral behavior.
1.3 Gender and Morality
There is no unanimous opinion in the literature about differences between genders in moral behavior and moral reasoning. In fact, the literature presents evidence for both similarities and differences between men and women’s moral judgment (e.g., Akaah, 1989; Fritzsche, 1997; Frankel et al., 1997; Ford & Richardson, 1994).
10
care for women (Lifton, 1982, p. 550). So, Gilligan showed that Kohlberg’s model is not applicable for women’s moral reasoning, because they have a different self-description model.
That said, there are a few critics of Gilligan’s model. For example, Friedman, Robinson and Friedman (1987) successfully disproved Gilligan’s theory. They found “highly similar patterns” of men’s and women’s moral judgments (Friedman et al., 1987, p. 44). The study was elaborated on the base of both Gilligan’s and Kohlberg’s ideas about moral development. Participants had to rank statements about moral dilemmas: half of the considerations were based on Gilligan’s view of women’s moral reasoning and half on Kohlberg’s description of principled stages. They also measured individuals’ sex attributes (Friedman et al., 1987). Findings of their research did not reveal any differences between male’s and female’s rankings. Measurement of personality also did not support the hypothesis about sex-differences in moral judgments (Friedman et al., 1987). Because of these potential differences in gender and moral behavior, I use gender as a control variable in my analyses.
11
II. METHODS
I collected this data as part of a larger experiment on moral and immoral behavior. Within this experiment, I collected individual-difference data on people’s LOC. In the current paper, I only discuss the part of the larger experiment that is relevant to my hypotheses.
2.1 Participants
The participants for this study were students of the Rijksuniversiteit Groningen (N = 81; 57% female) who volunteered to participate in an experiment for extra course credits or for material remuneration. Ages ranged from 18 to 30 years old (M = 21.7, SD = 2.75).
2.2 Design and Procedure
This study measured an individual-difference level trait (LOC) and how it was related to moral and immoral behavior. Following an introduction and signature of consent forms, participants were given the opportunity to show both immoral and moral behavior. They then completed the LOC scale and demographic measures.
2.3 Measures
12
2.3.2 Immoral behavior. I measured immoral behavior by giving participants the opportunity to cheat on a matrices task (Gino et al., 2009). Participants received a sheet with 20 matrices. Specifically, participants were asked identify the two cells in each matrix that added up to 10. They had 7 minutes to do so. After their 7 minutes were up, the experimenter came in to the room and told participants that they would receive €.20 for each matrix that they solved correctly. She gave the participant an answer key and an envelope with 20 coins. The experimenter told him that he should take out a coin for each matrix that he solved correctly, to leave the remainder in the envelope and to write on the outside of the envelope how many coins he kept. The envelope was then to be placed on the table. The experimenter also told the participant to insert his completed matrices sheet into a box outside of the lab room, because it was no longer needed. Thus, the participant thought that I was unable to track his actual performance. However, the materials were designed such that one of cells in the last matrix contained a number that matched the matrices sheet to the participant’s identification number. See Appendix A for the materials for this task. Thus, I measured immoral behavior as the extent of their actual cheating – that is, the difference between their actual and reported number of correct matrices.
13
III. RESULTS 3.1 Descriptive Statistics
Manipulation checks displayed that within 81 students, thirteen cheated (16%). For those who cheated, the mean number coins taken was 1.7 (SD = 1.68). For the moral behavior, 45 (55%) students donated their coupon to Greenpeace instead of using it for their personal use. Within men 63 % donated (n = 21) and 9% cheated (n = 3), whereas within females 50% donated (n =24) and 20% cheated (n = 10).
3.2 Hypotheses Testing
To examine Hypothesis 1, I performed a linear regression analysis predicting cheating from LOC. To examine Hypothesis 2, I used loglinear regression analysis to examine donation behavior from LOC. I also examined these hypotheses another way by not using LOC as a continuous variable, but rather to dichotomize the variable and look at differences between individuals with low and high LOC.
14 In sum, the results revealed that, taken together, the variables LOC, Coins Taken and Gender predict only 11% of the cheating results. Thus, I can assume the existence of other variables that predict cheating behavior, R2 = .108.
For further exploration of the results above, I divided the participants into two groups: external LOC and internal LOC, using the LOCSUM variable. The mean score of SUMLOC was a threshold value for the separation of participants into the two groups (M = 12.71). Thus, I considered people to have an external LOC if their LOC score was from 4 to 13 and I considered people to have an internal LOC if their LOC score was from 14 to 22. Dividing the sample in this way, a univariate ANOVA also did not reveal any significant relationship between these two groups, F(1, 79) = .180; p =. 672. That is, people with an internal LOC (M = .39, SD = .49) and people with an external LOC (M = .52, SD = .50) did not differ on their cheating behavior.
In a contrast to Hypothesis 2, a loglinear regression analysis of the data showed that in the current dataset the variables coupon (dependent variable), LOCSUM (predictive variable) and gender (control variable) are independent, meaning that there is no significant effect of participant’s LOC or gender on his behavior of donation. However, gender has marginally significant effect on donation. Table 1 depicts the effect of the variables, SUMLOC and gender on donation (coupon).
TABLE 1- coefficients for variables SUMLOC and Gender
15
IV. GENERAL DISCUSSION
In conclusion, the results of the analyses did not provide support for my hypotheses. There was no relation between LOC and cheating behavior and as such, I cannot conclude that people with an external LOC tend to behave more immorally than those with an internal LOC. In the same way, there was no significant relation between LOC and moral behavior (donation), which was predicted given that people with an external LOC tend to behave less morally than people with an internal LOC. In the following section, I elaborate on possible reasons for this null effect and why it differs from the results predicted in the existing literature.
In this study, I examine the relationship between LOC and moral behavior. In most current research examining the relationship between LOC and morality, the authors found a significant relationship, only when also looking at the presence of moderating factors, such as principled stages of CMD (Trevino & Youngblood, 1990), resistance to social pressure (Leftcourt, 1976), influence of outcomes expectancy (Trevino & Youngblood, 1990), and degree of field-dependence (Bloomerg & Soneson, 1976). Moreover, in each study moral or immoral behavior was measured in a different way: it could be the behavior of whistle – blowing (Dozier & Miceli, 1985) or helping behavior and altruism (Midlarski, 1971).
16
4.1 Hypothesis 1
The first hypothesis stated that external LOC is positively correlated with immoral behavior. The data showed that LOC has no link with a variable of cheating. I will elaborate on three particular reasons for it.
When taken as a whole, locus of control was not correlated with cheating, meaning that when a person feels in control of the outcomes of his actions he is not expected to behave more or less moral than those individuals who do not feel in control of their actions. One of the possible reasons for this null effect is that my study did not take into account any moderating factors, besides gender, and their influence on the dependent variable (cheating). Apparently there exist moderators that influence predictive ability of measured LOC on the moral/immoral behavior. Without including such moderators into the model, no correlation was found. This result is noteworthy, because researchers may not be able to judge person morality based on the LOC alone. Therefore, Rotter’s (1996) Internal/External LOC scale that could be used in HRM for recruitment purposes should not provide basis for judgments about subject’s morality.
Second, in the matrices task participants received remuneration for successful matrix combination totaling maximum of €4. This amount may have been an insignificant enough amount for people to cheat. Possibly, if the amount of reward was higher, and thus the payoff of immoral behavior increased, I would have found a different result.
4.2 Hypothesis 2
17 More than half (55%) of the participants opted to donate to Greenpeace instead of using the coupon for direct personal benefit independent of their LOC. Again, this can be explained in terms of the (not-measured) moderators such as: attitude toward the Greenpeace organization, their causes, attitudes towards donations, the amount of the donation, social pressure and so forth. If the reasons for internals to donate could be explained by their locus of control, externals might have engaged in donating behavior due to the presence of abovementioned moderators.
4.3 Limitations
This research has a number of limitations. First, this study was conducted in a laboratory setting; this means that the behavior was not it its natural setting. Even though students were informed that their participation was anonymous it cannot be stated that the same behavior would have been observed in real life. Second, it was conducted at a University. Attitudes towards accepted behavior at the University can prevent students from engaging natural behavior (i.e., that would have been observed in real life settings outside of the laboratory).
Another set of limitations concerns with the sample, namely: participants’ age, cultural diversity, occupation, and sample size. All participants were between 18-30 years old (M = 22 years old) and were primarily students of RUG University, possibly making my data narrow with results not representatively spread across the entire population. On the other hand, the sample included 10 ethnic groups from the sample size of 81. Therefore due to such high ethnic diversity of a relatively small sample specific cultural differences may play a too strong role in the final research outcomes.
18 And as I mentioned above, another possible limitation could be the relatively little payoff participants received for the experiment. They could have been inclined to engage in cheating if the material payoffs for such a behavior were significantly bigger.
Finally the research was conducted in English, which for a great number of students was not their native language, leaving the possibility for some participants to misunderstand or misinterpret the tasks.
4.4 Future Research
The current study could provide numerous avenues for future research.
An alternative variable for measuring moral behavior - donation to charity, particularly in this study - may provide further insights. Thus, relying on theoretical assumption that moral behavior is altruistically driven (School, 2008), the variable of moral behavior could be, for example, “offering a seat”.
This variable can be measured in the following laboratory conditions. A participant takes a seat in the waiting room where there is only one chair. After a while, a pregnant woman enters the room. Morality of behavior depends on whether the participant offers his/her chair to the woman or not. I could also involve an elderly or a physically challenged person to test moral behavior.
19 Payoff factor has proved to be important. Depending on its level people feel more or less inclined to cheat. It can be assumed that if the level is too low, the payoff can easily be neglected. On the other hand, if it is too high, personal and social pressure come into action causing low level of cheating as well. Such a hypothesis can be further elaborated in future research.
Reliability of findings increases in the context of moderating effects of individual attitudes towards various charity organizations and charitable causes. Donations to charities dealing with the disabled or ill children differ in occurrence with donations to pro-animal or anti-pollution organizations. This fact is worth to be analyzed in the future research.
Lastly research looking into differences of the LOC-moral/immoral behavior relationship across different cultures could provide interesting theoretical implications. For example one could hypothesize a correlation between the model of this research and cultural, economic, demographic, and social level of the country.
4.5 Practical Implications
Findings of the current research are likely to be used in business sphere, particular in the recruitment. Specifically, not experienced recruiters should take into account, that in selection process usage of simple techniques (LOC scale in this case) will not guarantee a forecasted behavior (moral/immoral). In the selection procedures of the recruitment processes the prediction of moral/immoral behavior should be based on other possible predictors, for example, resistance to social pressure (Leftcourt, 1976), influence of outcomes expectancy (Trevino & Youngblood, 1990), and degree of field-dependence (Bloomerg & Soneson, 1976), etc.
20 these individual’s differences, that role have to be further researched, particular its role in forecasting of moral/immoral behavior.
4.6 Conclusions
In the current research, I found no relationship between LOC and moral/immoral behavior. It is quite possible that a relationship does exist. However, without looking at moderating factors, it is quite difficult to actually find a relationship.
21
REFERENCES ARDdictionary.
http://ardictionary.com/Immoral/1006
Bandura, A. (1977). Social Learning Theory. New York: General Learning Press.
Bloomberg, M. and Soneson, S. (1976). The effects of locus of control and field
independence-dependence on moral reasoning.The Journal of Genetic Psychology, 128, 59-66.
Campbell, J. P. (1990). Modeling the performance prediction problem in industrial and organizational psychology. In M. D. Dunnette& L. M. Hough (Eds.), Handbook of Industrial and Organizational Psychology (pp. 687-732). PaloAlto, CA: ConsultingPsychologistsPress, Inc..
Crain, W. (1985). Chapter Seven. Kohlberg’s stages of moral development. Theories of Development. Prentice-Hall, 118-136.
Deigh, J. and Audi, R. (1995). The Cambridge Dictionary of Philosophy.
Dozier, J., and Miceli, M.P. (1985). Potential Predictors of Whistle-blowing: A Prosocial Perspective. Academy of Management Review, 10, 823-836.
Draper N, Smith H. (1998). Applied Regression Analysis. 3edition. ISBN: 978-0-471-170822.
Dunning, D. (2007). Self-image motives and consumer behavior: How sacrosanct self-beliefs sway preferences in the marketplace. Journal of Consumer Psychology, 17, 237–249.
Field, A. (2009). Discovering statistics using SPSS: and sex and drugs and rock ’n’roll. Thousand Oaks, California.
22
Business Ethics, 58, 65–77.
Friedman, W., Robinson, A. and Friedman, B. (1987). Sex differences in moral judgments? A test of Gilligan’s theory. Psychology of Women Quarterly, 11, 37-46.
Hodgkinson, Gerard P. (1992). Development and Validation of the Strategic Locus of Control Scale. Strategic Management Journal, 13(4), 311-317.
Hotelling, K., Forrest, L. (1985). Gilligan’s theory of sex-role development, a perspective for counseling. Journal of counseling and development, 64(3), 183-186.
Jordan, J., Mullen, E., and Murnighan, J. (2011). Striving for the Moral Self: The Effects of Recalling Past Moral Actions on Future Moral Behavior. Personality and Social Psychology Bulletin, 37(5), 701-713.
Jones, T. (1991). Ethical decision making by individuals in organizations: an issue-contingent model.Academy of Management Review, 16(2), 366-395.
Karabenick, S. and Srull, T. (1978). Effects of personality and situational variation in locus of control on cheating: Determinations of the “congruence effects”. Journal of Personality, 46, 72-95.
Kohlberg, L. (1969). Stages in the development of moral thought and action. New York: Holt, Rinehart & Winston.
Lefcourt, H.M. (1976). Locus of Control: Current Trends in Theory and Research. NJ: Lawrence Erlbaum Associates. ISBN 0-470-154044-0.
Lefrançois, G. (2000). Psychology for teaching: a bear is not a choirboy! Washington/Thomson Learning, Belmont, CA.
23
Assessment, 46(5), 550-552.
Midlarski, E. (1971). Aiding under stress: The effects of competence, dependency, visibility and fatalism. Journal of Personality, 39(1), 132-149.
Near, J. P., & Miceli, M. P. (1985) Organizational dissidence: The case of whistle-blowing. Journal of Business Ethics, 4, 1-16.
Pallant, J. (2005). SPSS survival manual: a step by step guide to data analysis using SPSS.Subon by Bookhouse, Sidney. ISBN 1 74114 478 7.
Oxford Dictionary of philosophy. (2008). 240.
Richmond, K.A. (2001). Ethical Reasoning, Machiavellian Behavior, and Gender: The Impact on Accounting Students’ Ethical Decision Making.
Rotter, J. B. (1966). Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs: General and Applied, SO, 1-28.
Rubin,S., Dierdorff, E., Brown, M. (2010).Do Ethical Leaders Get Ahead? Exploring Ethical Leadership and Promotability. Business Ethics Quarterly, 20(2), p. 215-236.
Scholl, R. W. (1981). An analysis of macro models of organizations: The goal and political models. Administration and Society, 13, 271-298.
Schulman, M. (2002). How we become moral. In C. R. Snyder, & S. J. Lopez (Eds.), Handbook of Positive Psychology (pp. 499-512). Oxford: University Press. Standler, R. (2004). Morality and Education.
http://www.rbs0.com/morality.pdf
24 of moral self-regulation. Psychological Science. 20(4) 523-528.
Trevino, L. K. (1986). Ethical Decision Making in Organizations: A Person-Situation Interactionist Model. Academy of Management Review, 11(3), 601-617.
Trevino, L. and Youngblood, S. ( 1990). Bad Apples in Bad Barrels: A Causal Analysis of Ethical Decision-Making Behavior. Journal of Applied Psychology, 75(4), 378-385.
Tsang, J. ( 2002). Moral Rationalization and the Integration of Situational Factors and
Psychological Processes in Immoral Behavior Review of General Psychology, 6(1), 25-50.
Varca, P. and James-Valutis, M. (1993). The Relationship of Ability and Satisfaction to Job Performance. Journal of Applied Psychology: An International Review, 42(3), 265-275.
Ward, L. and Wilson, J. (1980). Motivation and moral development as determinants of behavior of acquiescence and moral action. Journal of Social Psychology, 112, 271-286.
White, T. (1992). Business ethics and Carol Gilligan’s “two voices”. Business Ethics Quarterly, 2 (1), 51-61.
25 APPENDIX Matrices task Matrix Sheet 5.64 2.85 9.48 1.68 9.52 2.15 6.71 4.36 1.67 8.1 5.48 8.91 1.63 2.32 7.93 4.92 6.1 5.83 2.81 6.78 7.19 8.46 5.67 7.57 7.87 3.62 9.41 5.72 7.4 5.84 2.49 9.59 6.62 4.16 5.61 0.49 8.18 9.01 3.97 5.2 4.56 9.12 0.28 2.92 6.59 1.12 6.93 9.72 Found it Found it Found it Found it
1.43 2.11 5.36 7.45 8.57 9.3 5.39 2.29 0.42 3.28 4.43 2.6 3.99 3.91 4.34 8.39 2.72 0.51 9.61 3.57 2.36 6.09 4.56 1.58 4.74 4.23 1.34 2.1 5.49 4.21 6.26 7.86 9.78 0.89 2.14 6.71 6.58 8.97 3.21 3.85 3.87 5.82 4.14 9.46 6.15 9.12 4.2 6.83 Found it Found it Found it Found it
26 8.96 2.29 5.99 7.71 2.94 4.25 2.01 1.28 5.12 3.82 4.21 7.91 2.65 4.17 4.53 7.49 0.55 6.39 6.52 5.1 9.28 3.91 1.42 4.42 3.48 6.95 4.87 3.51 2.9 1.34 2.02 5.06 2.89 8.1 7.61 8.84 Found it Found it Found it Found it
3.53 5.94 4.16 6.15 2.97 2.32 1.9 9.3 7.68 3.55 4.72 2.32 0.74 4.55 3.19 8.17 7.91 1.83 5.62 0.81 2.15 3.75 3.72 2.09 1.08 5.19 6.6 6.56 3.06 2.72 8.98 9.22 3.39 0.71 3.44 4.48 0.17 7.13 7.2 2.54 5.16 5.49 9.48 8.5 9.71 2.87 6.86 1.23 Found it Found it Found it Found it
