Academic langauge

[Olga]

I was reading through research and found something not very ordernary regarding the langauge used in academic papers. The paper itself was boring and higly complicated to read. It was a research into cheating and moral behaviour. And suddenly they wrote something that was funny:

"Because the moral judgment of most adults in our society is at the conventional level (Colby et al. , 1983; Kohlberg, 1976), these results have implications for the functioning of our social institutions, many of which operate on the assumption of honesty and trust. Our findings suggest that those whose moral reasoning takes a societal perspective are indeed more likely to behave in accordance with society's rules. When behavior occurs in impersonal settings, such as a large classroom where cheating can occur, or a highway where traffic laws can be ignored, or a library where books can be removed without being checked out, or a subway where slugs can be inserted in token slots, Stage 3 subjects may not obey the rules because they apparently do not have the conception that the system will break down if everyone does not cooperate. It is sobering to realize that under the conditions of the present study, 96% of the subjects in the low-moral judgment group cheated, at least to some extent."


According to Kohlberg's theory the majority of people reason at this stage of moral reasoning - stage 3 - conventional. It is believed that criminals' moral reasoning is lower than of general populations.
The recent findings show that the majority of adult criminals also reason at this stage and some of the sexual offenders even on the higher stages.

[FDG]

Looks like kind of an obvious conclusion.

Anyway, in regard to the possibility that sexual offenders might show an higher level of moral behaviour. I would say that:

1)How is the sampling for evaluating the behaviour done?
- At every action of the indivudal has been given the same weight
- The actions are weighed

2)On how many sexual offenders?
3)How long is the interval of time?

Let's, for the sake of simplicity, just analyze the first two possibilities. If the sampling has been averaged without being weighed for the damage done by every action, then the proposition is meaningless, since it is clear that a single rape has much more weigh than cheating on an exam, or driving recklessly. I would go as far as saying that rape/murder/etc should have a +infinitum weight, so that a single murder would make the "moral ranking" of the raper immediatly fall to the lowest possible available class.

[FDG]

By the way, what does this have to do with academic language?

[Lewis Stern]

Most academic or scientifically objective works should appreciate the use of a language called e-prime in my view. For more information check out this site.

In a nutshell:
eStandard provokes 'heated' arguments because everyone generally writes in an arrogant tone. - "two wrongs don't make a right due to absolute conclusions."

ePrime creates 'warming' debates because everyone can have a true right to their opinion per se. "no right or wrong answer, only a relative conclusion." Combine say 10 relevant threads on a forum and we may approach an absolute conclusion for those who want to 'switch off' i.e. close-the-door or lid on a particular topic.

As far I understand most of us have been brought up in a world that naturally promotes Absolutism with a lack of regard for Relativism. When I first looked at this it opened-my-mind so to speak and now I believe I have a better balance in my writing, speaking and thinking style.

To cut a long story short, absolute statements, written, spoken, (thus thought of) can cause all kinds of problems with individuals. Relative, on the other hand, requires more integrated thinking with reality and opens the door to new discoveries in my view.

Ps. The use of 'is' is highly dangerous when a person remains ignorant of General Semantics and E-Prime in my view. Why? Because these subjects form the unsung basis for self-help programs.

As far as I know, the use of 'is' which equates to an '=' mathematical operator can 'program' a person's mind over time causing all sorts of problems, neuroses etc. The only people who should be doing any sort of neuro-linguistic programming should know what they're doing.

For anyone else, ignorance is dangerous!

With the use of the 'to be' verb anyone and everyone can be arrogant in their own style. Unless recognised they will continue to parade around provoking reactions in others in like 'a constant battle for intellectual surpemacy' in my perception.

Unlike the MBTI absolute 16 types 'boxing' system I don't believe that any one particular type 'is' arrogant. Although i do believe that some people love to believe that so they have a psuedo-reason i.e. 'the perfect rationalization' to bully those who they simply don't like based on a type that in reality means nothing, as Jung clearly stated.

Socionics types are relative. This is a key importance. I believe we've all done it i.e. 'boxed' ourselves and I applaud those who lose the seriousness of type and don't mention it in their signatures.

Knowing one's 'type' as Jung clearly states remains the intellectual property of the Practical Psychologist or in this case, a relative Socionist who probably knows what they know through open-minded thinking not closed-mind certifiable academia..

I believe this annoys people who want the satisfaction of feeling special with something to hold onto like someone who get's a degree and mistakenly believes their life is now complete.

Everything is relative otherwise everything only seems absolute at least until one 'retires' to the idea that no more can be learnt or invented. If believed then you remain 'in the box'.


When you think outside the box, the box goes away.

[Olga]

That was an intersting perspective, I did not think that far. I would agree with relativism. Before that I was not sure about it. It seems to me now like a good thing to have in mind.

I did not think far about sex offenders either. I just was reading the paper and it was really boring - no feelings no personal attachement so too speak - no colours! And when I have read what is expected from the the majority of populations - I thought - cool! We have not lost a sense of idiotism yet. Idiotism does not sound as absolutism to me but concepts are probably similar to a degree.

[olduser]

What is the method of testing?

[Lewis Stern]

A typical MBTI test (crudely) wants to 'box' you as follows:

Extrovert or Introvert

Sensing or Intuitive

Thinking or Feeling

Judging or Perceiving

The MBTI seems like a 'hip-hop style' dim-witted test i.e. E to the I, S to the I etc.

On the other hand, an atypical Socionics Type Test like the one here remains objective and scientific relative to the individual. You end up with an end-result except your 'box' is more like a 'box' with a dashed border to imply a) it's conceptual thus relative and b) it's not absolute.

I believe the only reason we can allow for subtypes is for those who really feel the need to narrow down and pin point who they 'are' further and can't remain content with a reasonable and relative conclusion.

If it's not relative then it's biased!

A general obsevation: Absolutism, and other -isms, lead to destructive arguements all in the name of 'being' right.

Relativism, a more scientific appreciation, requires 'harder', well-thought-out clear thinking to base an argument more on some thing said to drive progress not to fuel a heated debate for the sake of a thrill.

Given a choice I believe most people will always default on what seems 'easier' expecting miracles so to speak with what they think they know.

If one person is right then one person is wrong and if these persons come into contact then you have instant controversy. This way of thinking is the hallmark of Traditionalism in culture and education.

The 'cure' equates to relative and rational philosophies of education such as a combination of behaviourism and constructvism.

Do people adopt these (or understand them)?

Generally no because they would probably seem 'harder'.

And finally...

The paths of most resistance yield the most favouable results. The paths of least resistance yield the least favouable results.


So-called Absolutists want to re-iterate information over and over and over with minimal progression in an arguement. They usually hate a so-called 'smart alec' who adds a whole new line of thought. To keep you silenced they must use things like FEAR and INTIMIDATION to maintain the STATUS-QUO and their PSEUDO-POWER.


So-called Relativists want to build upon what has been said with any progression of an arguement. They usually receive flak from so-called 'Little ******s' who want to ban every new line of thought as heresy. To add a new line of thought will usually get met with AGITATION and DENIAL by those who can't accept that things (and people) do in fact CHANGE and IMPROVE.

[FDG]

There are so many wrong things and stereotypes that I do not know from where to start to break to pieces what you just said. Stay tuned, this post will be edited.

[MysticSonic]

I have to agree with FDG; Stren is so wrong that I don't know where to begin.

[Angel von Himmel]

A typical MBTI test (crudely) wants to 'box' you as follows:

Extrovert or Introvert

Sensing or Intuitive

Thinking or Feeling

Judging or Perceiving

The MBTI seems like a 'hip-hop style' dim-witted test i.e. E to the I, S to the I etc.

I think the main problem with this method is that it fails to address the issue of whether a person is using which introvert/extroverted functions, and then makes matters worse by blurring over the issues with unnecessary stereotyping. But, then again I have yet to see a test that successfully calculated types while using a method intended to deduct a type by gauging which introverted/extroverted functions a person uses, atleast I have personally seen minute success at this attempt on all tests I have made based on model-a.

[Lewis Stern]

Indeed.

Relative modelling may reach an absolute measurement of like 99.98% accuracy or whatever.

Enough for someone to feel comfortable in a 'box' even though there's like 00.02% possibility that in actual fact it all means nothing!

Hence leaving room for GROWTH and not STAGNATION with the 'perfect' rationalization of who you 'are'.

Finding your 'reference points' so to speak with such a degree of HIGH ACCURACY simply serves for the practical psychologist (inside of each and every one of us) to then map out the steps for change towards our goals, personal or otherwise.

If you're looking for an approval rating out of 10...

Do you want a final ten out of ten or a nine out of ten implying always room for improvement?

Those who wish to 'switch off' and think inside the box can gain a short-lived 'high' feeling like a mr or mrs "perfect" with their 10.

Given a choice I would gladly strive for 1st place (as is the competitve nature of business) though fully appreciate that everyone must get knocked off their 'high horse' thus remain no.2 for a modest time.

I once remember getting accused of being a perfectionist. I now believe that perfectionism is more prevalent then you think and cannot simply be attributed to a particular type. It would seem wise to accept that perfection is a moving target i.e. relative.

[Olga]

I did not understand what is it exactly you see wrong with Lewis post.
My senses did not detect any crime so far.

@Heath. I will pm you with more info about the test or tell more here when I finish with my piece of work.

[Olga]

I think I should have put it not here - it is not about socionics. But becasue a few people were interested...

Moral Reasoning and Moral Conduct : An Investigation Prompted by Kohlberg's Theory


By: Carl I. Malinowski
City University of New York Graduate School

Charles P. Smith
City University of New York Graduate School


This report is based, in part, on a dissertation by Malinowski (1979). The authors wish to thank Ian D. McMahan, Herbert Saltzstein, and Jeffrey Shaw for helpful suggestions, and Gavin Schneider for serving as research assistant. We also wish to acknowledge the many valuable suggestions made by two anonymous reviewers. Carl I. Malinowski is now at Pace University.


The revival of interest in moral education reflects a concern not merely to encourage more sophisticated moral thinking but ultimately to foster an increase in moral behavior as well (Kohlberg, 1981, p. 302). But is it a safe assumption that greater maturity of moral thinking will lead to a higher degree of moral behavior? According to Kohlberg (1969), maturity of moral judgment "can be a quite powerful and meaningful predictor of action" (p. 397). However, a recent review of relevant research reveals mixed results and concludes that cognitive developmental theory "offers only the vaguest guidelines for approaching the relations of cognition and action" (Blasi, 1980, p. 1).

We, too, find Kohlberg's statements regarding moral cognition and moral action to be both incomplete and imprecise. However, we also believe that his position, as stated, has not received adequate investigation, due both to methodological limitations of prior research and to the omission in previous studies of a number of relevant factors affecting the relation between moral judgment and behavior. Consequently, we have attempted first to pull together Kohlberg's main propositions regarding the relation of moral reasoning and behavior, and second to investigate some of their implications.

In the present research we selected resistance to temptation as the moral behavior of interest, and investigated the determinants of cheating in an achievement-oriented testing situation. The distinctive features of the present study include its approach to the measurement of moral judgment, the development of a sensitive measure of the degree and latency of cheating that is suitable for adults, the assessment of factors affecting the strength of temptation, the specification of Person × Situation interactions, and a focus on behavioral differences among adult subjects at Kohlberg's two conventional stages of moral reasoning.


Kohlberg's Views on Cognition and Action

In his cognitive-developmental theory, Kohlberg (1981) postulates an invariant sequence of stages of moral development beginning with obedience to external rules and fear of punishment and culminating in adherence to self-chosen universal principles of justice. In the interest of brevity, we have attempted to summarize in outline form the major aspects of that portion of Kohlberg's theory that deals with the relation of moral judgment to moral behavior.

1. Moral judgment determines action by way of concrete definitions of rights and duties in a situation (Kohlberg, 1971, p. 229). For example, at the conventional level a person might refrain from cheating in order to maintain the good opinion of a person in authority (Stage 3), or because it is believed that the social system would break down if everyone cheated (Stage 4) (Kohlberg, 1976).

2. The greater one's maturity of moral judgment, the more one is predisposed to behave morally (Kohlberg, 1969; Kohlberg & Colby, 1983).

• As moral maturity increases, there is an increased sense of personal responsibility to carry out actions regarded as morally correct (Candee & Kohlberg, 1982; Kohlberg & Colby, 1983).


• The influence on behavior of moral affect (e. g. , guilt) depends on the level of moral judgment (Kohlberg & Colby, 1983). For example, "two adolescents, thinking of stealing, may have the same feeling of anxiety in the pit of their stomachs. One adolescent (Stage 2) interprets the feeling as ‘being chicken’ and ignores it. The other (Stage 4) interprets the feeling as ‘the warning of my conscience’ and decides accordingly" (Kohlberg, 1971, pp. 189-190).


3. Moral behavior cannot be predicted from moral judgment alone.

• A person can know what is right and not do it.


One cannot follow moral principles ... if one does not understand or believe in them. One can, however, reason in terms of such principles and not live up to them. A variety of factors determines whether a particular person will live up to his stage of moral reasoning in a particular situation (Kohlberg, 1976, p. 32).



• The personality determinants of moral behavior include both moral and nonmoral factors. The former include moral judgment and moral affects, for example, guilt, empathy, (Kohlberg, 1969, p. 390). For Kohlberg the relevant nonmoral factors are "ego strength" factors such as the ability to control impulses and delay of gratification (Candee & Kohlberg, 1982).


• Situational factors affect the relation between moral maturity and moral behavior. For example, Stage 3 individuals might refrain from competitive cheating against a friend but not against a stranger. Principled individuals might decide to cheat if other considerations make cheating appear to be a relatively moral act (Kohlberg, 1969).


• The effect of situational factors on behavior is moderated by moral judgment. "The higher the individual's moral stage, the more able is he to resist situational forces toward behavior inconsistent with his moral judgment structures" (Kohlberg & Freundlich, 1975, p. 15).



Other Views of Moral Judgment and Behavior

Rest (1979) also provides a thoughtful discussion of the relation of moral judgment to behavior. He emphasizes, and Blasi (1980) would agree, that "a moral judgment test only gives us information about a subject's concepts of fairness; however, many other psychological processes are involved in interpreting situations and organizing one's actions" (p. 179).

Norma Haan (1978) believes that Kohlberg's stages of formal reasoning are more likely to be used for hypothetical dilemmas than for real action situations. She regards Kohlberg's theory as incomplete in that other factors besides formal reasoning, namely, interpersonal and contextual considerations, affect moral action. (See also Bloom, 1977; Gilligan, 1982. )

Our own approach develops more fully the role of nonmoral motivational influences on moral decisions (cf. Schwartz, Feldman, Brown, & Heingartner, 1969; Smith, Ryan, & Diggins, 1972). We believe that an individual at the conventional level of moral reasoning will yield to temptation when the strength of temptation exceeds the strength of moral restraint. However, the strength of temptation may not be the same for all subjects given the same objective incentive to cheat (e. g. , the opportunity to win a monetary prize). In our view, the strength of temptation is determined not only by the nature of the incentive, but also by the person's need or motive for that incentive. Thus, the prediction of resistance to temptation involves an assessment of both strength of temptation and of factors affecting moral restraint.

In the present experiment, as an incentive to falsify performance scores, we caused the subject to attain embarrassingly low practice scores. Through extensive pretesting we determined the strength of a failure induction that would cause at least 50% of the subjects to cheat. We also measured individual differences in the strength of two motives relevant to the avoidance of failure and humiliation, namely, test anxiety and need for approval. We anticipated a greater strength of temptation, and hence a greater incidence of cheating, from those in whom these motives were strongest.


Prior Research

In a recent review, Blasi (1980) identified 11 published and unpublished studies relating moral judgment (as measured by the Kohlberg interview) to cheating. One additional study of this type not mentioned by Blasi is by Fodor (1972). Of these 12 studies, 6 show the expected positive relation between moral judgment and resistance to temptation, 4 fail to show such a relation, and 2 report mixed results. All but two of these studies (viz. , Gallagher, 1975, and Schwartz et al. , 1969) used children or adolescents, or both, as subjects—groups that do not permit consideration of the full range of moral maturity. Moreover, the fact that fifth and sixth grade children are classified in Stages 5 and 6 in several of these studies raises questions about the appropriateness of the moral judgment scoring, because Kohlberg (1976) now feels that no one reaches the principled level of moral reasoning until late adolescence. In fact, in terms of Kohlberg's most recent scoring system (Colby, Kohlberg, Gibbs, & Lieberman, 1983), individuals do not even reach consistent Stage 4 reasoning until they are in their 20s and 30s (cf. Fischer, 1983).

Two additional studies used the Defining Issues Test (Rest, Cooper, Coder, Masanz, & Anderson, 1974) as a measure of moral judgment with college students as subjects. Dunivant (1976) reported a correlation of . 18 (p < . 05) between moral judgment and resistance to temptation, and Leming (1978) found the expected relation between moral judgment and cheating in only one of two conditions.

On the whole, these 14 studies provide modest support for Kohlberg's expectations regarding moral judgment and moral conduct, but it also seems clear that the conditions under which moral judgment predicts resistance to temptation are not fully understood. In addition to the reservations concerning this body of research expressed by Blasi (1980), we note that few studies assessed the strength of the temptation that was aroused by the introduction of incentives to cheat, and none of the studies predicted or found a significant difference in cheating between Stage 3 and Stage 4 subjects. Because most adolescents and adults are at these stages (Kohlberg, 1976), we believe it is a matter of considerable practical and theoretical interest whether Kohlberg's theory predicts differences in behavior between stages as well as between levels.

In our view, a statisfactory study of cheating normally requires: (a) a situation in which there is a low perceived risk of detection, (b) a nonobvious measure of cheating that is sensitive to different degrees and latencies of cheating, (c) specification of the incentives to cheat, (d) assessment of the degree of temptation to cheat, (e) assessment of subjects' suspicion regarding the true purpose of the research, and (f) a careful debriefing.


Major Hypotheses

Our summary of Kohlberg's views leads to our main hypothesis:


H1: The higher the moral maturity, the lower the incidence of cheating, and the greater the latency of cheating.


(Our summary also indicates that an inverse relation between moral judgment and dishonesty is not expected under all conditions, but it should occur when there is an implicit or explicit understanding that cheating is not to take place, and when there are not other considerations that make cheating appear to be a relatively moral act. )

More specifically, we also hypothesize that in normative testing situations, Stage 3 subjects are more likely than Stage 4 subjects to cheat, and to begin cheating sooner.

Although we believe this more specific hypothesis follows from Kohlberg's theory, Kohlberg himself generally speaks as though he expects differences in cheating between levels of moral judgment (preconventional, conventional, postconventional) but not between stages within levels (Kohlberg, 1969, p. 396; see also Schwartz, et al. , 1969). For example, Kohlberg (1971) says, "conventional subjects cheat not because their restraint of impulse is less than that of principled subjects, but because their cognitive definition of right and wrong is less independent of what other people think" (p. 231).

According to Kohlberg's writings, however, the morality of Stage 4 subjects is more internalized than that of Stage 3 subjects. Stage 4 subjects see more clearly the ramifications of an act for the entire social system, and have a greater sense of obligation to fulfill duties to which they have agreed (Kohlberg, 1976). For Stage 3 subjects, the perspective is interpersonal rather than societal. Knowledge that "everyone is doing it" should have a greater effect on Stage 3 subjects in whom "there is much conformity to stereotypical images of what is majority or ‘natural’ behavior" (Kohlberg, 1981, p. 18).

Kohlberg also implies that guilt is more likely to be a deterrent to dishonesty for Stage 4 than for Stage 3 subjects. He says, "Stage 4 guilt implies differentiating concern about one's responsibility according to rules from Stage 3 ‘shame’ or concern about the diffuse disapproval of others" (Kohlberg, 1969, p. 391).

For all these reasons we expect a difference in the degree and latency of cheating between Stages 3 and 4. These same considerations also lead to the following hypothesis:


H2: Guilt concerning transgression should be a stronger deterrent to cheating in Stage 4 than in Stage 3 subjects.


According to Kohlberg, consensus and interpersonal concordance form the basis of morality at Stage 3. Knowledge that a previous subject had cheated should have a disinhibiting effect when Stage 3 individuals are deciding "whether to abide by existing norms or the common expectations of ... [the] group" (Rest, 1974, p. 65). On the other hand, consensus that cheating is necessary to "succeed" at a task should not lead Stage 4 subjects to violate the cultural prohibition against cheating. Cheating would still be wrong because of a "commitment to pre-existing rules and expectations" (Kohlberg, 1971, p. 202). Accordingly, we hypothesize that:


H3: The introduction of a confederate who says that he cheated should increase cheating more for Stage 3 than for Stage 4 subjects.



Subsidiary Hypotheses

As Zimmerman (1983) has pointed out: "Nearly all cognitive theories accept the premise that prediction of actual performance requires consideration of motivational processes" (p. 20). We regard test anxiety as indicative of the strength of motivation to avoid failure. Because cheating on a test serves as added protection against failure, subjects with high-test anxiety should be more inclined to cheat than those with low-test anxiety (cf. Hill & Eaton, 1977; Smith, Ryan, & Diggins, 1972, among other studies that demonstrate this relation). We therefore hypothesize that:


H4a: Under testlike conditions, the higher the test anxiety, the greater the incidence of cheating and the sooner cheating will begin.


Kohlberg's (1969) belief that nonmoral affective factors influence moral decisions less as the stage of moral development becomes higher suggests an additional hypothesis:


H4b: The relation of test anxiety to cheating will be stronger for Stage 3 than for Stage 4 subjects.


We follow Jacobson, Berger, and Millham (1970) in assuming that the Social Desirability Scale (Crowne & Marlowe, 1964) measures a motive to avoid social disapproval, and that such an avoidance motive will be aroused in situations in which a subject confronts failure that might lead to humiliation or rejection. Consequently, we hypothesize:


H5a: Under conditions in which failure is likely, the higher the need for approval, the greater the incidence of cheating and the sooner cheating will begin.



H5b: The relation of need for approval to cheating will be stronger for Stage 3 than for Stage 4 subjects.



Method

Overview

Level of moral judgment was assessed in a group testing situation. The remaining measures were obtained later in apparently unrelated individual testing sessions. Subjects were randomly assigned to one of two conditions, a control condition in which a confederate, posing as another subject, said nothing about cheating, or an experimental condition in which the confederate told the subject that he had cheated. Hypotheses are tested by correlations (using one-tailed p values) and by analyses of variance (ANOVA s) comparing dependent variable scores of subjects high versus low in moral judgment.

Subjects

The participants were 53 male college students, mainly freshmen and sophomores, who were selected in the following manner: First, a research assistant administered the Defining Issues Test (Rest et al. , 1974) during class sessions of introductory-level social science courses. Later, the senior author recruited males whose Defining Issues Test responses were complete and consistent for an apparently unrelated study of "attention-concentration. " Subjects were offered $2. 50 for participation and were told that all data would be anonymous.

Measures

Level of moral judgment was measured by the Defining Issues Test, which presents six hypothetical moral dilemmas, each followed by 12 statements exemplifying Kohlberg's different stages of moral reasoning. Subjects rank the four issues they regard as most important in reasoning about each dilemma. Continuous scores are obtained reflecting Rest's (1979) view that "in general, principled thinking gradually increases over the course of development" (p. 101). Relative to the Kohlberg interview, the major advantages of the Defining Issues Test are group administration and objective scoring.

The selection of this measure, and the means of scoring used, were based on the results of a pilot study in which 32 male introductory psychology students were given the Defining Issues Test and later the Kohlberg moral judgment interview (Form A). The Kohlberg interview responses were scored using the structural issue scoring system by the senior author who had attended a scoring workshop at Harvard University. K-20 scores, based on the highest stage at which at least 20% of a subject's responses are scored, were obtained. A subset of 11 protocols was scored by another Harvard-trained scorer. 1 Interscorer agreement was 91%.

The K-20 scores were correlated with two types of scores obtained from the Defining Issues Test: (a) a principled morality or "P" score indicating degree of preference for statements exemplifying Stages 5 and 6 (r = . 34, p < . 06, N = 32) and (b) the sum of weighted ranks given to stage 4 and "A" (antiestablishment) items added to the P score, making a "4AP" score (r = . 66, p < . 001, N = . 32). 2 Because the 4AP scores were most nearly equivalent to the scores obtained from the Kohlberg interview, it was decided that in the present study, 4AP scores would be used in addition to the customary P scores.

According to the K-20 scores, 11 of the pilot subjects were at Stage 4, 18 at Stage 3, and 1 at Stage 2. This distribution suggests that most of the students in the college population used would be classified as Stage 3 or Stage 4 by the Kohlberg interview.

Although the correlation between the K-20 scores and the 4AP scores is only a modest . 66, there is a more impressive relationship between the 4AP scores and moral stage as determined by the Kohlberg interview. Of the subjects classified as Stage 4, 10 out of 11 had 4AP scores above 66. 7; of those classified as Stage 3 or below, 18 out of 21 had 4AP scores of 66. 7 or below χ2 (1,N = 32) = 14. 53, p < . 001.

This cut-off point provided a basis for classifying subjects in the present study as Stage 4 and above or as Stage 3 and below. A 4AP score of 66. 7 was used as the dividing line between subjects considered relatively high (n) = 30) or low (n = 23) in moral judgment. Because of the correspondence between 4AP scores and the Kohlberg stages, the high group was assumed to consist predominantly of Stage 4 subjects, and the low group predominantly of Stage 3 subjects. Thus, in the analysis of data relating moral judgment to cheating, this high-low division makes possible an approximate comparison between Stages 3 and 4.

Guilt was measured by means by five self-report items that were intended to assess different aspects of guilt. For each item, extent of guilt was indicated on a 9-point scale. One item was taken from MacKinnon (1938); the others were devised for this study. The measure was pretested, and intercorrelations among the items revealed two fairly distinct clusters. Two items composd an anticipatory guilt cluster: "To what extent do you actually feel guilty when considering doing something wrong?" and "To what extent are you able to anticipate feeling guilty when you consider doing something wrong?" The other three items composed a posttransgression guilt cluster: "Do you, in everyday life, often feel guilty about things which you have done or have not done" (MacKinnon, 1938). "How many different things that you do or do not do make you feel guilty in the course of everyday living?" "Compared to most other people, how guilt-prone a person would you say you are?" Because the scale items were not homogeneous, it was decided to obtain separate scores for anticipatory and for posttransgressional guilt.

Test anxiety was measured by means of Part I of the Test Anxiety Questionnaire (Mandler & Sarason, 1952). Scores from Part I have been shown to correlate between . 84 and . 90 with scores from the entire Test Anxiety Questionnaire (Smith, 1965).

Need for approval was measured by means of the Social Desirability Scale (Crowne & Marlowe, 1964).

Cheating was measured by determining the extent to which a subject reported better scores than he actually made in an achievement-oriented testing situation. Extensive pretesting was conducted to create a situation in which the perceived risk of detection was low and the temptation to cheat high. After several tasks were tried and rejected, a rotary pursuit task was devised that was highly involving and did not arouse suspicion. The nonacademic and nonverbal nature of the task had the further advantage that the personality characteristics affecting performance were (presumably) not the same as those affecting scores on the Defining Issues Test.

The apparatus used was a Polar Pursuit Tracker, Model PR-15 (Research Media, Inc. , Syosset, NY). The task called for holding a stylus over a light that moved along a triangular route at 30 rotations per min. After a series of practice trials, the subject was left alone to take 10 "test" trials of 10 s each. The "time on target" the subject recorded for each trial was compared with the true time read from a clock in another room. Three measures of cheating were obtained: (a) the number of trials on which cheating occurred (trials cheated), (b) total time, in seconds, falsely reported over the 10 trials (seconds cheated), and (c) the first trial on which cheating occurred (latency of cheating).

Degree of temptation to cheat was assessed in a post-debriefing questionnaire by means of the following question, accompanied by a scale ranging from 0 to 100: "Indicate the strength of the incentive or temptation to cheat in the situation you were just in, as you experienced the situation. "

Procedure

When the subject arrived at the laboratory, he was left alone briefly with a confederate who was introduced as "the previous subject. " Subjects were randomly assigned to a control (n = 28) or an experimental (n = 25) group. To those in the experimental group, but not to those in the control group, the confederate gave the impression that it would be necessary to cheat on the performance task in order to get a decent score.

After the confederate left, the experimenter, who did not know whether a subject was in the experimental or the control condition, returned and introduced the rotary pursuit as a measure of attention-concentration, and gave instructions that emphasized the importance of good performance for success in certain occupations (e. g. , astronaut). Subjects were urged to do their best, and were given false and unattainably high norms purporting to be those for athletes (highest), college students, school children, and institutionalized children (lowest). A ticking metronome provided a background distraction that lent credibility to the notion that concentration was being studied.

A sheet was provided for recording the time for each trial. It included the initials and (fictitious) scores of "previous subjects. " The subject was asked to put his initials at the top of the column in which he would record his times. After five practice trials the experimenter informed the subject that he would need to make rapid improvement in order to do as well as other students. Then the experimenter left the room on the pretext of starting another subject. In order to reduce further the perceived risk of detection, the subject was told to practice a while, then take his 10 test trials, and then come to the experimenter's office. When the subject appeared with his scores, he was given the need for approval, test anxiety, and guilt measures.

Before the formal debriefing began, the subject's level of suspiciousness was ascertained by asking him some questions about the cover story including the following: "Was there ever a time in which you thought something other than attention-concentration was being assessed" (and if so, what)? If he revealed any suspiciousness at all that cheating was being studied, he was asked to estimate (from 0 to 100%) how strongly he suspected this. Then he was fully debriefed, reassured of anonymity, and told that he could either take his data and withdraw, or remove his initials, turn in the data, and sign a consent form allowing use of his data. All subjects agreed to have their data used. In approving this research, the Institutional Review Board specified the use of this consent procedure and an especially thorough debriefing. 3


Results

Effectiveness of the Manipulations

The results indicate that the procedures effectively aroused temptation because 77% of the subjects cheated on at least one trial. The item assessing degree of temptation is significantly related to trials cheated (r = . 45, p < . 001), seconds cheated (r = . 39, p < . 01) and latency (r = −. 38, p < . 01). (Unless otherwise noted, N = 53, and p values are two-tailed. )

Suspiciousness scores are not significantly correlated with any of the three measures of cheating. Of the 53 subjects, 41 said they had no suspiciousness at all. Of the remainder, 7 said they were more than 50% suspicious. Because analyses including and excluding these 7 subjects are virtually identical, only the analyses involving all 53 subjects are reported in detail.

Moral Judgment and Cheating

According to Hypothesis 1, the higher the moral judgment scores, the less the cheating and the greater the latency of cheating.



Table 1 presents the correlations between both indexes of moral judgment and each of the other variables. The P (principled morality) and 4AP scores are correlated . 73. Because they are somewhat stronger, only correlations between the P scores and the other variables will be cited in the text.

As expected, moral judgment is negatively related to the number of trials on which subjects cheated and the number of seconds by which subjects inflated their scores. P scores are correlated −. 48 (p < . 001, one-tailed) with trials cheated, and −. 39 (p < . 002, one-tailed) with seconds cheated. Also as expected, moral judgment is positively related to latency. The higher the P score, the later a subject began to cheat (r = . 43, p < . 001, one-tailed). All these correlations remain significant (p < . 005) when suspiciousness scores are partialed out.

It was also predicted that in comparison with Stage 4 subjects, Stage 3 subjects will cheat more and begin cheating sooner. The results clearly support this expectation. A comparison of the number who cheat or do not cheat in the high versus low moral judgment groups yields a Yates' corrected χ2 (1,N = 53) of 6. 03 p < . 02). It is especially noteworthy that of the 12 persons who did not cheat, 11 are classified as high in moral judgment.

Cheating scores of subjects high or low in moral judgment in the control or experimental conditions were compared by means of 2 × 2 ANOVA s. For each measure of cheating, a main effect for moral judgment is significant. As compared with subjects high in moral judgment, those low in moral judgment cheated on significantly more trials, F(1,49) = 11. 08, p < . 002, recorded significantly higher seconds cheated, F(1,49) = 4. 59, p < . 04, and started cheating sooner, F(1,49) = 10. 94, p < . 002. Thus, Hypothesis 1 receives strong support both from the correlational analysis using P scores and from the high-low split based on the 4AP scores.

Guilt and Cheating

Table 1 indicates that although post-transgressional guilt is not related to cheating, anticipatory guilt is significantly related to trials cheated (r = −. 26, p < . 05, one-tailed) and to seconds cheated (r = −. 23, p < . 05, one-tailed) in the expected direction.

Because moral judgment and anticipatory guilt were found to be independent (see Table 1), cheating scores of subjects high or low in moral judgment and high or low in anticipatory guilt were compared by means of 2 × 2 ANOVA s. The main effects agree with the correlational analysis in showing that anticipatory guilt is significantly related to cheating for all subjects combined: For trials cheated F(1,49) = 10. 28, p < . 002; for seconds cheated F(1,49) = 11. 91, p < . 001. However, Hypothesis 2 is not supported, because there are no significant interaction effects. That is, anticipatory guilt is not a stronger deterrent to cheating for Stage 4 than for Stage 3 subjects.

Confederate Conditions and Cheating

Hypothesis 3 states that the presence of another subject who says he cheated should increase cheating more for Stage 3 than for Stage 4 subjects. ANOVA s comparing cheating scores for subjects high or low in moral judgment in the control or experimental conditions reveal no significant main effects due to conditions, and no significant interactions. These results are not altered by an analysis in which suspiciousness scores are covaried. Thus, Hypothesis 3 is not supported.

Test Anxiety and Cheating

Hypotheses 4a and 4b state that subjects with high test anxiety are expected to cheat more often and sooner than those with low test anxiety, and that this difference should be more pronounced for Stage 3 than for Stage 4 subjects. ANOVA s with high-low moral judgment and high-low test anxiety as the factors reveal a significant main effect for test anxiety on trials cheated, F(1,49) = 4. 56, p < . 04, and on latency, F(1,49) = 4. 02, p < . 05. These results support Hypothesis 4a. However, the results do not support Hypothesis 4b, as there are no interactions indicating that the expected relation is stronger for the low than for the high moral judgment group.

Need for Approval and Cheating

Hypothesis 5a states that a positive relation is expected between the need for approval and cheating. ANOVA s with high-low moral judgment and high-low need for approval as factors reveal a main effect for only one of the three measures of cheating; namely latency, F(1,49) = 4. 62, p < . 05. That is, subjects above the median in need for approval begin to cheat significantly sooner than those with scores below the median. This result provides partial support for Hypothesis 5a. However, there are no interactions indicating that the expected relation is stronger for Stage 3 than for Stage 4 subjects. Thus, hypothesis 5b is not supported.

Practice Trial Performance and Cheating

One might expect that those subjects whose practice trial times were best would be least tempted to cheat. In fact, this appears to be the case. The correlations in Table 1 indicate that the better the practice trial performance, the fewer the number of trials cheated (p < . 001), the fewer the number of seconds cheated (p < . 001), and the later the cheating began (p < . 003).

When practice trial time, which is correlated . 18 with moral judgment, is partialed out, the correlations remain significant (p < . 005, one-tailed) between P scores and all three measures of cheating. Thus, the relation between moral judgment and cheating is not simply due to the fact that subjects with higher moral judgment scores also score higher on the performance task.

Finally, it is of interest to ask to what extent the independent variables combine to predict cheating. The multiple correlations of the independent variables (P score, practice time, anticipatory guilt, test anxiety, and need for approval) with the dependent variables are as follows: (a) with trials cheated, R = . 64, (b) with seconds cheated, R = . 61, (c) with latency, R = . 58. By themselves P scores and practice time contribute, about equally, almost 90% of the variance accounted for. Anticipatory guilt, test anxiety, and need for approval make minor additional contributions.


Discussion

Implications for Kohlberg's Theory

The present study deals only with that part of Kohlberg's theory that concerns the relation of moral judgment to moral conduct. We have not attempted to test developmental hypotheses; hence, we assume rather than demonstrate that Stage 4 reasoning is more mature than Stage 3 reasoning. Moreover, our measure of moral reasoning is continuous rather than discrete. Like Rest (1979) and Fischer (1983) among others, we do not subscribe to a "structured whole" view of stages, because subjects' moral reasoning about various issues is seldom confined to a single stage, and because changes in moral reasoning tend to be gradual rather than saltatory. Nevertheless, we believe that there are different types of moral reasoning, and we are persuaded by Kohlberg's evidence that, at least for males, they tend to emerge in a developmental sequence (Colby et al. , 1983).

We should also note that our results were obtained with male subjects and may or may not generalize to female subjects. The research of Haan (1978) and Gilligan (1982) among others, suggests that females are more affected than males by the extent to which a situation involves a subject's responsibility for other persons.

Whether persons whose moral reasoning is more advanced behave in a more moral way than those whose reasoning is less advanced is a matter of considerable theoretical and practical importance. Kohlberg clearly specified that such a relation is expected under the appropriate conditions, though he does not specify in detail what those conditions are. Our aim was to investigate some of the implications of this view with particular attention to the role of situational factors and nonmoral personality characteristics.

The major hypothesis received clear support, namely, the higher the moral judgment scores, the less the cheating, and the later the cheating occurred. Because the measures of moral judgment and cheating were quite dissimilar, it seems unlikely that the relation between them is due to method variance.

Because the measure of moral judgment is continuous, these results do not necessitate strong assumptions about stages. In the most general sense they show that different types of moral reasoning are associated with different degrees of resistance to temptation in an achievement-related context. However, the present research was initiated partly because Kohlberg's propositions also lead one to expect differences in moral behavior among persons within moral judgment levels as well as between levels. Using Kohlberg interview scores from our pilot study data we derived a cut-off point for our 4AP scores that classifies subjects as predominantly Stage 4 or predominantly Stage 3. Comparisons using this procedure provide strong support for the hypothesis that under conditions in which cheating is implicitly prohibited, Stage 4 subjects will resist the temptation to cheat to a greater extent than Stage 3 subjects.

Because the moral judgment of most adults in our society is at the conventional level (Colby et al. , 1983; Kohlberg, 1976), these results have implications for the functioning of our social institutions, many of which operate on the assumption of honesty and trust. Our findings suggest that those whose moral reasoning takes a societal perspective are indeed more likely to behave in accordance with society's rules. When behavior occurs in impersonal settings, such as a large classroom where cheating can occur, or a highway where traffic laws can be ignored, or a library where books can be removed without being checked out, or a subway where slugs can be inserted in token slots, Stage 3 subjects may not obey the rules because they apparently do not have the conception that the system will break down if everyone does not cooperate. It is sobering to realize that under the conditions of the present study, 96% of the subjects in the low-moral judgment group cheated, at least to some extent.

However, it would be a mistake to overstate the difference between Stage 3 and Stage 4 subjects. Although substantially fewer Stage 4 subjects cheated, a majority of them did eventually cheat to some extent. On the average, Stage 3 subjects who cheated began to do so on the second trial; Stage 4 subjects who cheated began to do so on the third trial. Our interpretation is that temptation to cheat (in order to avoid embarrassingly low scores) increased as low scores accumulated, and that moral restraint was eventually overridden by temptation for all but 12 of the 53 subjects.

Although our data reveal the expected relation between moral reasoning and cheating, they do not confirm our hypothesized interactions. The relation to cheating of anticipatory guilt, test anxiety, and need for approval is not significantly different for Stage 3 than for Stage 4 subjects. Kohlberg might see these results as consistent with the conception of a qualitative shift between the conventional and the principled levels. He says that "Affective-situational forces are less determining of moral decisions at the principled than at the conventional level... At the principled level ... resistance to temptation is less contingent on ‘strength of will’ ..." (Kohlberg, 1969, p. 369). Perhaps attaining the level of self-chosen principles places one above temptation. If that is the case, one would expect the predicted interactions to occur only when comparing subjects at the principled level with those below that level.

Knowledge that another cheated.

Learning that another subject had cheated was expected to disinhibit Stage 3 subjects more than Stage 4 subjects. This hypothesis was not supported, but for reasons that have nothing to do with whether the theory is right or wrong. In order to show the predicted effect, more Stage 3 subjects in the experimental group would have to cheat than Stage 3 subjects in the control group. While every Stage 3 subject in the experimental group did cheat, all but one of the Stage 3 subjects in the control group also cheated, making the baseline rate of cheating too high to permit an increase due to the role of the confederate. Hence, this hypothesis, which we continue to regard as viable, will have to await a more adequate test.

Nonmoral Personality Factors and Cheating

As Blasi (1980) notes, "How noncognitive motivations interact with moral reasoning as facilitators or inhibitors of moral behavior has been rarely studied..." (p. 10). In the present study it was assumed that the arousal of two nonmoral motives, test anxiety, and need for approval, would create a desire to avoid the humiliation and embarrassment of failure and thereby increase the temptation to report falsely high scores. As noted earlier, the results confirmed these expectations to some degree. Both motives were significantly related to latency of cheating, and test anxiety was also significantly related to the number of trials on which a subject cheated. Perhaps the relationship of need for approval to cheating would have been stronger if the situation had entailed a greater possibility of social disapproval or rejection as a result of poor performance.

As Schwartz et al. (1969) point out, the relevant nonmoral factors operating in a situation may affect the relation between moral judgment and behavior; or as Staub (1982) has put it, how moral orientation is expressed in behavior depends on what other goals are activated. Motives may be aroused that produce behavioral tendencies consistent with a person's moral inclinations as, for example, when achievement motivation aroused in a testing situation strengthens the tendency not to cheat (Schwartz et al. , 1969; Smith et al. , 1972). Alternatively, the arousal of motivation such as the desire to avoid failure may augment temptation to cheat and, if strong enough, override moral reservations against cheating, at least for subjects at the conventional level.

This point of view has important implications for whether moral behavior, or any other behavior, is expected to be consistent across situations. Persons bring to different situations certain relatively stable dispositions, such as characteristic moral reasoning, ego strength, and motives. The behavior that will occur in a particular situation depends on the strength and variety of such dispositions, the extent to which these dispositions are aroused by the perceived situation, and the manner in which the resulting behavioral tendencies interact to produce behavior. Only a theory that specifies the relevant person-situation determinants of behavior can predict the extent to which behavior will be similar across situations (cf. Schwartz et al. , 1969).


Conclusion

After years of focusing primarily on thinking, cognitive developmentalists are again turning their attention to the relation of their research to behavior. In the present study we have attempted to contribute to the empirical literature on moral thought and action by testing hypotheses about adult behavior under theoretically relevant conditions in which both amount and latency of cheating are measured. As expected on the basis of Kohlberg's theory, higher scores on the Defining Issues Test are associated with lower cheating and greater latency of cheating. Also, subjects classified as predominantly Stage 3 are more likely to cheat than those classified as predominantly Stage 4, though the latter may also cheat if temptation becomes strong enough.

Although the results pertaining to our main hypothesis are consistent with Kohlberg's theory, they do not preclude alternative interpretations. For example, if different moral points of view are associated with different personological structures (cf. Johnson, Hogan, Zonderman, Collins, & Rogolsky, 1981), then it may be the correlated aspects of character (e. g. , independence, trust) that account for different degrees of resistance to temptation rather than the stage of moral reasoning per se.

However, it must be acknowledged that it was Kohlberg's identification of the different kinds of reasoning assessed here that led to our hypotheses. Not all of these hypotheses were supported, however. Although anticipatory guilt is related to cheating, it is not correlated with moral judgment, as might be expected, and it is not a greater deterrent to cheating for Stage 4 than for Stage 3 subjects. Neither are the predicted interactions obtained for the relation of test anxiety and need for approval to cheating.

To put the results into perspective, we call attention to the relative effects on cheating of situational factors, moral reasoning, moral affect, and nonmoral motives. As the pattern of correlations in Table 1 indicates, the strongest determinants of cheating in our data are P scores and situational factors related to strength of temptation (viz. , temptation, practice time). Next in order of strength comes moral affect (anticipatory guilt), and finally, nonmoral motives. Thus, it is clear that moral judgment is only one of several determinants of behavior which, in our data, is largely independent of the others. We believe that further progress in understanding the determinants of moral behavior will depend on the development of theories that are as complex as the phenomena to be understood—theories that reflect the effects of social and contextual factors as well as gender and nonmoral personality characteristics.

[Lewis Stern]

@olga

just a little tip:

reading long posts on a forum can be very difficult on the eyes and brain i.e. it doen't read well or in the case of socionics, it doesn't get absorbed/metabolised effectively in my opinion.

as far i know the recommend width of an article should be around 50-75 characters and around 750 words long like i use on my site.

i'm not saying i'm not interested in the above research, just making a point that it hasn't engaged me (yet!).

[Creepy-pokeball]

Hey Olga. I have a huge migraine (no more parties...please ) but I will read it later. I do remember, though, that the last moral stage was pretty much impossble. I need to find it again...

[FDG]

In a nutshell:
eStandard provokes 'heated' arguments because everyone generally writes in an arrogant tone. - "two wrongs don't make a right due to absolute conclusions."

Are you sure you aren't projecting your own arrogance into the statements of other peole? How are you not arrogan in proposing a language which eliminates arrogance? How is your supporting position less arrogant than a position which holds "eStandard" true?

ePrime creates 'warming' debates because everyone can have a true right to their opinion per se.

Everybody has the right to his opinion right now, too. I like it, because this way I can demolish crap like this.

"no right or wrong answer, only a relative conclusion."

You're saying that the right answer is that there is only a relative conclusion. Congratulations, here your degree in coherence of thought. Not.

Combine say 10 relevant threads on a forum and we may approach an absolute conclusion for those who want to 'switch off' i.e. close-the-door or lid on a particular topic.

How do you define how many threads are enough? Do you relize that even a post of yours can be infectious to a whole collection of topic so that the expected relevance goes to zero in the interval of "snowyc post [1,+infinitum]?

As far I understand most of us have been brought up in a world that naturally promotes Absolutism with a lack of regard for Relativism. When I first looked at this it opened-my-mind so to speak and now I believe I have a better balance in my writing, speaking and thinking style.

I've already disproved above your relativist nonsense. Relativists hold true that truths are only relative. However, since this proposition serves as axiom, it is necessarily absolutely true for your relativistic framework to function. Therefore, the founding axiom contradicts itself.
Gratz, it's not easy to do.

As far as I know, the use of 'is' which equates to an '=' mathematical operator can 'program' a person's mind over time causing all sorts of problems, neuroses etc. The only people who should be doing any sort of neuro-linguistic programming should know what they're doing.

I have already disproved that your relativistic framework does not require the usage of the identity operator.

Unlike the MBTI absolute 16 types 'boxing' system I don't believe that any one particular type 'is' arrogant. Although i do believe that some people love to believe that so they have a psuedo-reason i.e. 'the perfect rationalization' to bully those who they simply don't like based on a type that in reality means nothing, as Jung clearly stated.

No, the basis is idiotic beliefs held by the subject. Which is the case with you, my dear.

The rest you say is rethorical crap not worthy of my time.[/img]

[Creepy-pokeball]

Here ya go Olga

[Olga]

Thanks Jadae, it was a good article and I saved it. I am about to finish and to send my MSc dissertation which is about moral reasoning, personality and nature of offence. Lets hope and pray for me from the 14 July onwards. All I want is just to pass really, study is hard work and I would rather prefer to deal with socionics. I feel that I had enough of official studying.

I would not advice to read that long research I posted - I should not have done it. I did though use the findings of Malinowski in my piece of work. Well, it was already established that moral reasoning is not a major determinant of offending behaviour. The recent studies are digging for other variables or interaction of moral and non-moral factors mediating moral reasoning. Aleixo and Norris pointed towards psychoticism and general intellectual ability, Addad was studying personality differences and meaning of life. My mini-research found significant difference between violent and non-violent on life value. It turned out that violent offenders appreciate life more that non-violent. I had to explain why it may be the case on the basis of existing research evidence. I wonder what other forum members are studying or doing professionally - I mean what are their interests about?