Orientation Observation In-depth interviews Document analysis and semiology Conversation and discourse analysis Secondary Data Surveys Experiments Ethics Research outcomes



Social Research Glossary

About Researching the Real World



© Lee Harvey 2012–2020

Page updated 29 April, 2020

Citation reference: Harvey, L., 2012–2020, Researching the Real World, available at
All rights belong to author.


A Guide to Methodology

CASE STUDY Authoritarian Personality

The Authoritarian Personality (Adorno et al., 1950) was a piece of critical social research directed at the psychology of prejudice. It is often seen as one of the primary Critical theory studies of the Frankfurt School. It was one of a series of interrelated studies undertaken by the Department Of Social Research of the American Jewish Committee and entrusted to the reconstituted Institue For Social Research in New York (formerly the Institut fur Socialforschung in Frankfurt [The Frankfurt School]). These studies were to deal with psychological and social features of prejudice.

Broad 'major hypothesis':

the economic and social convictions of an individual often form a broad and coherent pattern, as if bound together by a "mentality" or "spirit", and that this pattern is an expression of deep-lying trends in his personality.

The researchers major concern was with what they called the 'potentially fascistic individual'. This they regarded as somone who is particularly susceptible to anti-democratic propoganda. This susceptibility, they supposed, was a feature of the individuals's 'structure'. 

There is a lot of confusion amongst critics of the Authoritarian Personality about how the cental concept Authoritarian Personality actually emerged; fuelled by the late explanation in the published report. Was it stumbled upon as a result of empirical enquiry or had it been theoretically predefined? Was it construct of the study or a pre-existing concept of radical social philosophy. What the Authoritarian Personality failed to do was to clearly show that the concept had emerged dialectically.

The approach to the study

The Authoritarian Personality addressed prejudice through a conception of three levels of personality:

  • surface opinions, attitudes and values;
  • ideological trends indirectly manifested;
  • personality forces of the unconscious.

The authors wanted to reveal all three levels and so used a mixture of depth interviews (almost psychoanalytic interviews) and questionnaire test items.

The research was not a study of a population to see if it exibited anti-democratic tendencies as a whole but a comparison of two subsets, those that did and those that did not exhibit anti-democratic tendencies.

The research evolved; there was not a single research instrument that was, devised, administered and then analysed.

A questionnaire was used to select subjects for depth interviews and the interviews used to refine the questionnaire. A reciprocal development of research tools.

The questionnaire contained 'factual' questions of a broad classificatory type and Likert-scale type attitude questions. The latter were used to 'obtain quantitative estimates of certain surface ideological trends'. Four scales resulted: Ant-Semitism (A-S), Ethnocentrism (E), Politico-economic Conservativism (PEC ), Fascism (F).

The Anti-Semitism (A-S) scale, for example, was composed of 52 questions administered in two blocks, of 26 items, a week apart. The A-S consisted of five sub-scales labelled 'threatening', 'offensive', 'seclusive', 'intrusive' and 'action-attitudes'. Half of each scale was administered on each occasion. The scale items were based on the writings and discussions of US and European anti-semites and their 'rational' opponents.

The scale items were scored from 1 to 7 (least to most anti-semitic) no middle point score of 4 allowed. (Total score was from 52 to 364 for the 52-point scale). Statistical anlysis of a pilot survey of 144 women students on each sub-scale showed a good range of scores and good split-half reliability with Spearman-Brown coefficient of 0.89.  The intercorrelation between sub-scales was very high (0.92 to 0.94). This led to the pragmatic conclusion that the 52-point item was a useful measure of anti-semitism. Item analysis was used to 'weed out' items with low discriminatory power and with some rephrasing a shortened scale with ten items was devised. The 10-point A-S scale showed high split-half reliability coefficients (.89-.94) and an average DP of 3.68.

This new short scale was not simply the result of statistical analysis of the original items. It was also contingent upon the emerging thesis about the nature of anti-semitism that the authors were developing.

The use of the A-S scale alone on a sample of students at George Washington University showed to the researchers that (in this case) proximity to a substantial Jewish community increased prejudice amongst anti-semites rather than reduced it. Essentially, attempts to overcome prejudicial ideology by substituting facts that 'rationally' attack prejudice is inadequate as prejudicial ideology is resistant to factual correction. Thus they argued anti-semitic ideology is an all-inclusive standpoint that goes beyond the issue of racial minorities. Essentially, they argued that anti-Semitic views were related not to 'The Jewish problem' but to a whole range of [structural] traits including stereotyping, middle-class values, moral-purity of the in-group, immorality of the outgroup, fear of outgroup power, and  contamination, infiltration and victimisation by out-group.

So the next step was to examine anti-semitism in the context of prejudice in general to see if the same trends were evident in other ideological areas or in 'non-deological thinking' [From a critical social research perspective, of course, there can be no such thing as non-ideological thinking, hence the inverted commas].

The same process was used for the E, PEC and F scales.

The theoretical 'basis' of the E scale (14 items - 3 subscales) is that ethnocentrism is based on pervasive and rigid ingroup-outgroup distinction involving streotyped negative imagery and hostile attitudes towards outgroups and the converse for ingroups. Additionally, a hierarchical, authoritairan view of group interaction in which ingroups are dominant and outgroups subordinate is held.

This they related to the wider issues of politico-economic conservatism (via the PEC scale) in which the authors concluded that ethnocentrism is one aspect of a broader 'pattern of social thinking and group patterning'. They thus expected to find a central and 'subideological' psychological disposition from which all observed ideologies stem.

This was the basis for the F scale which (in retrospect) seems to be the goal to which they were aiming from the start. This was to be used instead of A-S and E scales (not PEC as did not corrrelate well). It was to be less obvious, i.e., an indirect measure of prejudice. Sources for the items were the A-S, E and PEC scales, available literature, previous research, results of Projective Questions and Thematic Apperception tests and the results of the depth interviews.

The authors specified nine characteristics of the 'prejudiced authoritarian individual'.

1. Conventionalism—rigid adherence to conventional, middle-class values. (6 items)
2. Authoritarian submission—submissive uncritical attitude toward idealised moral authorities of the ingroup. (8)
3. Authoritarian aggression—tendency to be on the lookout for, condemn, reject and punish people who violate conventional values (5)
4. Anti-intraception—opposition to the subjective, the imaginative, the tender-minded. (6)
5. Superstition and stereotypy—the belief in mystical determinants of the individual's fate, the disposition to think in rigid categories. (5)
6. Preoccupation with power and 'toughness'—concern with the dominance-submission, strong-weak, leader-follower dimension; identification with power figures; overemphasis upon the conventionalized attributes of the ego; exaggerated assertion of strength and toughness. (5)
7. Destructiveness and cynicism—generalised hostility, vilification of the human. (11)
8. Projectivity—disposition to believe that wild and dangerous things go on in the world, the projection outward of unconscious emotional impulses. (5)
9. Sex—exaggerated concern with sexual activity. (4)

This was a total of 55 items although only 38 different as 17 items were used twice on different sub-scales.

Although not explained, presumably the number of items in each section represented the theoretical importance of each section.

The meaning of an individuals response is a function of the entire pattern of responses not individual items responses.

A consolodated 78-item scale (Form78) resulted which was administered in 1945 to 4 groups of subjects. Indirect measures showed a reduced split half reliability (0.74) and only 3 of the 9 subscales had an average DP in excess of 2.

The test items rather than the underlying theory was seen to be at fault and a new shorted version was developed (Form 45 and Form 40). The F scale was not however shown to be unidimensional although it did 'hang together'. The reduced E scale was shown to be unideminsional. Overall the scale and its statistical analysis were somewhat suspect.

The responses were of no interest in themselves, they were indicators of fundamental differences in personality traits. The empirical information was interwoven with the theoretical approach.

Should the authors, nonetheless, have been more rigorous in their development, testing and analysis of the test items? Possibly, but not if it meant losing the 'wood for the trees'. It was important to keep a holistic analysis and a conceptual one—not get bogged down in statistics and reliablity measures, which are conventional positivistic concerns of 'surface analysis'. Get beneath the surface. However, this doesn't mean be shoddy, unsystematic etc, nor getting away with ignorance.

The clinical interviews were depth interviews but there was perceived to be a problem of 'objective and systematic' interpretation. The interviewers had a checklist of things to cover: vocation, income, religion, family background, childhood, sex, social relationships, school, politics, minorities and race (with some 'Suggested direct questions'.) A systematic set of categories (around 90) operating at three levels: factual; attitudinal; interpretive/technical clinical. In all, 80 usable protocols were rated. The results coincided with the attitude scales.

Use of projective questions (what would you do if you had 6 months to live...) and thematic apperception tests (what's happening in this ambiguous picture) were criticised (especially by positivists) as lacking objectivity. Other objections included:

1. Sample skewed towards young, West Coast, middle class, educated.
2. Only negative items on the scales.
3. Volunter respondents in the main.
4. Other variables, e.g., education, group membership ignored.
5. The apparent stereotype answers are a result of the scales that were essentially too unrefined.
6. The F scale had two incompatible goals: viz testing relation of authoritarianism to prejudice and measuring prejudice indirectly. The pragmatic determination of test items undermines the theoretical analysis.
7. Quest and clinical interviews were not independent and so cannot claim one confirmed the other [Although this was not, of course, the main concern]
8. Use of 'irrationality' explanations based on psychoanalytic categories to explain away apparent contradictions. I.e., the objection is (to the way) that the authors go beyond surface appearances.
9. Non-sociological, so nothing on group determinants of attitudes.

Practical conclusions

Counter-measures designed to combat prejudice should take into account the whole structure of the prejudiced outlook and so concentrate not on particular minority discrimination but on phenomena like stereotyping, emotional coldness, identification with power, general destructiveness. Rational argument and appeals for sympathy are futile against the prejudicial person and the best way of neutralizing prejudice is the use of the established characteristics of the prejudiced person - i.e. conventionality and submissiveness towards authoprity - including legal restraints. BUt this is no cure as the formatuion of the prejudiced personality is underpinned by child-parent relations (which is also related to socio-economic processes) and occurs at an early age. This cannot be headed off by psychologists alone and needs to be attacked by all social scientists.

The Authoritarian Personality and critical social research

As noted, a conventional criticism of the Authoritarian Personality has been that the researchers simply 'discovered' the authoritarian personality syndrome that they had already constructed.

This is, of course, a serious accusation. It is also a problem that all critical social research is liable to face. Why?

Principally, because critical social research does not take surface appearances for granted. The alternative conception is regarded by conventional critics as 'subjective' while their own taken-for-granted reification of appearances is supposed to be 'objective'. Quite clearly critical social research has to challenge the objective-subjective distinction from the outset. Further, it is quite clear also that critical social research has to legitimate its knowledge and not simply establish it, unlike positivistic social science.

Ironically, with the Authoritarian Personality we have a piece of critical social research that bends over backwards to accommodate the empiricism of positivistic research practices and is criticised for pre-ordaining its outcome.

How is critical social research to react to this problem ?

Arguably it is necessary to underake an extensive theoretical analysis (both historical and structural) of existing social structures and prevailing taken-for-granted conceptualisations. The deconstruction process takes place and the root for reconstruction outlined. Most important it must emphasise that this is a dialectical process (of theory and fact) and one in which the 'fundamental unit' or other reconstructive principle emerges necessarily from the ongoing analysis and critique of theory and empirical information. I.e. this is not a linear process.

So the critical social research perspective has to be established—painstakingly from first principles if necessary (e.g. Marx's ananysis of commodities) and, arguably, must include an elaboration of the oppressive mechanism. (Authoritarian Personality did not really do this in its report until rather too late; near the end and then pretended it was empirically/inductively derived and not dialectical).

This case study draws on Madge 1963


Return to Critical approaches and surveys (Section 8.2.7)