Make your own free website on Tripod.com
**************************************************************************************

Personality & Individual Differences, Vol. 1, pp. 303 to 305, 1980

ACQUIESCENCE AND THE WILSON CONSERVATISM SCALE



J. J. RAY

University of New South Wales

(Received 3 December 1979)

Summary -- Data contributed by six different researchers in Australia, South Africa, New Zealand and England were assembled from seven different administrations of the Wilson Conservatism scale. The correlation between the positive and negative halves of the scale was found to be generally around -0.5. It was concluded that although the C-scale is seriously affected by acquiescent response set, the balanced structure of the scale controls for this quite satisfactorily.


INTRODUCTION

As is perhaps inevitable in a widely-used instrument, the Wilson Conservatism scale (1973) has attracted criticism from several sources. Not all users have been equally satisfied with the measure it provides. One issue that has been raised on several occasions (Ray, 1971; Ray and Pratt, 1979) is the extent to which responses on the scale are affected by acquiescence. Wilson (1973) believes the scale to be exceptionally free from the influence of acquiescent response set but other results have shown it to be seriously affected.

Wilson (1974) however claims that the occasions on which the correlation between the positive and negative halves of the scale were low are exceptional. Where Ray and Pratt (1979) report correiations as low as -0.2, Wilson claims correlations of around -0.7. Where the lower figure is found, the scale would appear to be seriously lacking in construct validity. The available published data, however, do not allow prospective users to decide which correlation is most likely to be found on any given occasion. There are too few cases to decide whether the Wilson result or the Ray result is most common.

The present paper, therefore, is an attempt to find out what result is generally being obtained by other users of the scale. The answer should tell us whether or not the influence of acquiescent response set is so severe that the Wilson scale must be considered unusable in the purpose for which it was designed.

METHOD

The published literature using the Wilson scale was surveyed and all authors of papers using the scale (other than Wilson and his London associates) who could be found were circularized with a request for a copy of their raw data. People known personally to the author as users of the scale were also approached.

The response was rather disappointing. Many authors claimed to have lost data and others did not reply. The data that was received was analysed in terms of the pre-reversal correlation between the halves of the scale and in terms of the reliability of the scale with and without reversals. The latter statistic is a direct measure of the extent to which acquiescence rather than some substantive construct accounts for the responses to the scale items (cf. Martin, 1964). Reliability in both cases was assessed by use of Cronbach's (1951) coefficient "alpha".

RESULTS

The correlations and reliabilities observed can be read off from Table 1. It will be seen that neither the -0.7 result claimed by Wilson nor the -0.2 result reported by Ray and Pratt (1979) was in fact characteristic. Instead these two results define extremes on a continuum across which other results range. Generally, however, the results listed are eminently satisfactory. They offer strong support for the construct validity of the scale. The supposedly opposed items were in fact responded to in generally highly opposite ways.

It is also of some note, however, that the reliability of the scale scored for acquiescence is also very high. The levels observed were in fact quite similar to those reported in Ray and Pratt (1979). This means that acquiescence is elicited by C-scale items at an exceptionally high level. The C-scale does produce a consistent tendency in people to acquiesce.

DISCUSSION

The results obtained may at first sight appear to contain an element of contradiction. Does not one index show the C-scale to be substantially affected by acquiescence while the other does not? In fact both indexes do show an effect due to acquiescence. The average split-half correlation of the scale in the Jamieson data, for instance, may be calculated from alpha (See Ray, 1979) as being 0.75 [alpha/(2-alpha)]. This compares with 0.46 for the correlation observed between the two particular halves defined by the positively-scored and negatively-scored items. As this split-half correlation is the one most affected by acquiescence, one can see that acquiescence has in fact depressed the correlation concerned below what it should be if no acquiescence were present. A correlation of 0.46 is much less than a correlation of 0.75.

The point of importance for prospective C-scale users, however, is that the balanced structure and large number of items in the C-scale have successfully prevented the acquiescence from having any systematically distorting effect on C-scale final scores. Any balanced scale would of course secure this result but only a well-constructed scale would show meaningful correlations between the oppositely-scored halves. Wilson's strategy of using a large number of items for each of his halves has then succeeded in aggregating in each a degree of meaningful response which is sufficient substantially to overcome the effects of acquiescence. AIthough acquiescence causes all items to be responded to similarly, Wilson's items have enough meaningful content for respondents to respond to the two different types of item in highly opposed ways.

Thus, although it must be concluded that the C-scale can generally be used with confidence in its psychometric properties, it should be remembered that some occasions do arise when its psychometric properties indicate a substantial lack of construct validity. Although such occasions are rare, users would be well advised to examine the correlation between the two halves of the scale whenever possible.

Table 1. C-scale statistics observed on seven different samples. See Jamieson (1978), Barling and Fincham (1979), Barling and Evans (1980), Baker (1979), Kirton (1978). No separate report is available on the Brown or Pearson data.

Source.........Pos-neg correl.....alpha with revs....alpha without revs............N.....Sample

L. B. Brown....... -0.52.......................0.85...................0.51......................47....New Zealand Univ. students
D. N. Baker.......-0.37........................0.85...................0.67....................119....Australian Univ. students
B. D. Jamieson.. -0.46.......................0.86...................0.60....................168.....New Zealand community
M. J. Kirton....... -0.57.......................0.87..................0.48.....................208.....English general population
J. Barling.......... -0.45.........................0.87..................0.64.....................313.....S. African Univ. students
J. Barling.......... -0.55.........................0.82..................0.46.......................55.....S. African Univ. students
P.R. Pearson..... -.41.........................0.82...................0.55......................75.....English students

Acknowledgements -- The author would like to thank Ms D. N. Baker of the N.S.W. Health Commission, Prof. L. B. Brown of the University of N.S.W. School of Psychology, Dr. J. Barling of the Dept. of Psychology at the University of the Witwatersrand, Dr. M. J. Kirton of the Hatfield Polytechnic Occupational Research Centre, Dr. B. D. Jamieson of the Dept. of Psychology at the University of Canterbury and Mr. P. R. Pearson of the Psychology Dept. of the Pastures Hospital, Mickleover, Derby, who alone made this study possible.

REFERENCES


{Articles below by J.J. Ray can generally be accessed simply by clicking on the name of the article. I am however also gradually putting online a lot of abstracts, extracts and summaries from older articles by other authors so if an article not highlighted below seems of particular interest, clicking here or here might just save you a trip to the library}

BAKER D. N. (1979) Rape: Attitudes which facilitate defensive attribution towards the victim and attacker. Unpublished Master's dissertation, School of Psychology, University of N.S.W.

BARLING J. and EVANS I. M. (1980) Multidimensional psychological conservatism in South Africa: The case of an authoritarian culture. Unpublished :

BARLING J. and FINCHAM F. (1979) The effects of alcohol on psychological conservatism. J. soc. Psychol. 107, 129-130.

CRONBACH, L. J. (1951) Coefficient alpha and the internal structure of tests. Psychometrika 16, 297-334.

JAMIESON B. D. (1978) Scaling conservatism: The C-scale and Conscale II. Unpublished.

KIRTON M. J. (1978) Wilson and Patterson's conservatism scale: A shortened alternative form. Br. J. soc. clin. Psychol. 17, 319-323.

MARTIN J. (1964) Acquiescence-Measurement and theory. Br. J. soc: clin. Psychol. 3, 216-225.

RAY, J.J. (1971) "A new measure of conservatism" -- Its limitations. British Journal of Social & Clinical Psychology, 10, 79-80.

RAY, J.J. (1979) Is the acquiescent response style not so mythical after all? Some results from a successful balanced F scale. Journal of Personality Assessment 43, 638-643.

RAY, J.J. & PRATT, G.J. (1979) Is the influence of acquiescence on "catchphrase" type attitude scale items not so mythical after all? Australian Journal of Psychology 31, 73-78.

WILSON G. D. (1973) The Psychology of Conservatism. Academic Press, London.

WILSON G. D. (1974) Evaluation of the conservatism scale: A reply to Ray. N.Z. Psychologist 3, 27.



POST-PUBLICATION ADDENDA

1. The correlation between original C-scale halves reported in Ray & Pratt (1979) was -0.199 for 110 Army conscripts. As this implies that the two halves of the scale had less than 4% of their variance in common, it shows that the two halves of the scale were on that occasion essentially unrelated rather than measuring the same thing.
2. That completely anomalous correlations (such as the significant positive correlation of 0.288 between supposedly liberal and conservative items reported in Ray [1972]) can also be obtained, even with students, does mean that the many studies which continue not to report the correlation between the "liberal" and "conservative" items continue to be of dubious meaning. Accepting their results at face value is a matter of faith rather than of science.
3. This article should be read in conjunction with: Ray, J.J. (1985) Acquiescent response bias as a recurrent psychometric disease: Conservatism in Japan, the U.S.A. and New Zealand. Psychologische Beitraege 27, 113-119.

Reference:

RAY, J.J. (1972) Are conservatism scales irreversible? British J. Social & Clinical Psychology 11, 346-352.






Go to Index page for this site

Go to John Ray's "Tongue Tied" blog (Backup here)
Go to John Ray's "Dissecting Leftism" blog (Backup here)
Go to John Ray's "Australian Politics" blog (Backup here)
Go to John Ray's "Gun Watch" blog (Backup here)
Go to John Ray's "Education Watch" blog (Backup here)
Go to John Ray's "Socialized Medicine" blog (Backup here)
Go to John Ray's "Political Correctness Watch" blog (Backup here)
Go to John Ray's "Greenie Watch" blog (Backup here)
Go to John Ray's "Leftists as Elitists" blog (Not now regularly updated)
Go to John Ray's "Marx & Engels in their own words" blog (Not now regularly updated)
Go to John Ray's "A scripture blog" (Not now regularly updated)
Go to John Ray's recipe blog (Not now regularly updated -- Backup here)

Go to John Ray's Main academic menu
Go to Menu of recent writings
Go to John Ray's basic home page
Go to John Ray's pictorial Home Page (Backup here)
Go to Selected pictures from John Ray's blogs (Backup here)
Go to Another picture page (Best with broadband)