A primary objective of the personality study was the desire to provide a more satisfactory instrument which is based on scientific method and which can be used in the selection of OR scientists. With this perspective in view we now discuss how the 16PF, an instrument based on scientific foundation can assist in the selection of OR scientists.
An important and basic assumption in profile-matching for determining an individual's suitability to enter a particular occupation is that the mean profile of people who have stayed long in the particular occupation and who are generally believed to be satisfying the demands of that occupation and who are also believed to have adjusted to the particular demands of that occupation represent in a sense an "ideal" pattern for selecting new entrants into that particular occupation. (Cattell, Eber and Tatsuoka 1970). That is not to say of course, that the people presently engaged in the particular occupation whose mean profile is used as a guide present the best profile for effectiveness in the particular occupation. The profile is in error, only to the extent that the biases and prejudices, and limitations of the prevailing methods of recruitment into the particular occupation failed to attract "better" individuals into that occupation. There are basically two major approaches to profile-matching (Cattell, Eber and Tatsuoka 197U).
(i) Type Placement Method. Under this method, which is based on discrimi nant techniques, the profile of the prospective entrant to the
particular occupation is matched on the sixteen personality factors to test for the entrants "belongingness" to, or "adjustment" in the particular occupation; in other words, is the candidate a "member" of that group?
(ii) The Effectiveness Method. In this method, which is based on linear (or multiple) regression techniques, the sten scores of the candidate on the sixteen personality factors are multiplied by appropriate loadings or weights on each of the sixteen personality factors in order to test for the probability of sucess in the particular occ upa tion.
In the following sections, Type Placement Method is used to describe an approach to profile comparison. By comparing an individual's 16PF profile with that of the "standard", the individual's belongingness or otherwise to the group can be t ested objectively. This comparison can be done using coefficient of pattern similarity, rp (Cattell 1949). Pattern analysis is a complicated problem in profile comparison and there are a number of methods suggested for tackling the problem (Guión 1965). Mosel and Roberts
(1954) after comparing several methods concluded that "Cattell's (1949) coefficient of pattern similarity, rp , agreed best with commonsense judgements" (Guión 1965, p.175).
This method of profile comparison would be discussed in the following sections.
Another method of comparing 16PF profiles is that presented by Krug (1981). In Krug's approach, the second-order scores derived from the 16PF primary scores are calculated on the four factors: extraversión, anxiety, tough poise, and independence. The scores on these four second-order factors are then converted into a four-digit pattern code using some rules which are provided. The pattern code is then compared with a corresponding one from among the 81 patterns provided by Krug in order to get the interpre tation of the particular profile.
"represents a reasonably exhaustive sample of most variations in 16PF profiles one is likely to encounter in practice." (p.22).
Since the 1BPF profile of OR scientists is not among the occupational profiles provided in Krug the method is not directly relevant to the purposes of this research. We shall now discuss profile comparison based on Type Placement Method using pattern similarity coefficient, rp .
Comparison of 16PF Profiles:
Pattern similarity coefficient rp , presents mathematico-statistical properties useful for grouping people into types, and for making a number of comparisons of personality profiles such as individual-to-individual comparison, individual-to-group comparison, and group-to-group comparison.
(Cattell 1949, Horn 1961, Cattell and coutter 1966). As Cattell et al (1970) pointed out:
"...for many situations, the pattern similarity coefficient is the best thing available; for, in general, a person can be considered well
adjusted to a job in proportion to his resemblance to those in it" (p.144)
Individual-to-Group Comparison of 16PF Profile:
Pattern similarity coefficient rp , is perhaps the most widely used technique in adjustment placement of candidates into occupations. It allows the personality profile of the individual to be compared with the mean profile of the occupation.
By computing the individual's pattern similarity coefficient, tests can be conducted to determine how significantly the individual's profile resembles that of the group. An important advantage of this approach is that it tests for the adjustment of the "whole" person rather than for
his similarity on a few factors of personality
The pattern similarity coefficient is calculated in an individual-to-group comparison by using the formula (Cattell et al 1970, p.141):
rp = (4K + Z D 2 ) - £ d 2 (4K + Z D 2 ) + Z d 2 where:
d = difference in sten score on each factor between the individual's score and the group mean,
D = difference between the group mean sten score on each factor and the general population mean, 5*5,
K = Median of the chi-squared distribution with degrees of freedom equal to the number of personality factors measured (sixteen in the 16 P F ) The pattern similarity coefficient rp , takes values varying from +1 to -1 with "0" indicating no resemblance. The significance of rp can be evaluated using Horn's table (Horn 1961).
In interpreting rp values, it should be understood that the higher the value, the stronger the resemblance.
Group-to-Group Comparison of 16PF Profile:
Later on in this chapter, the profiles of five occupations which have some affinity with OR would be compared on a factor by factor basis using Student's t-test. As useful as such a comparison is, it has the d i s advantage that it tends to portray a "mechanistic" picture of the o ccupa tion's profile in that it emphasises similarities and differences on
particular factors rather than on the occupation's profile as a whole unit. An interesting comparison is the group-to-group comparison. In this case the occupation's profile is treated as a homogeneous group, that is, the mean profile can be thought of as the profile of an "ideal'* person representing the particular occupation.
In a group-to-group comparison, rp is calculated using the formula (Cattell et al 1970, p.311)j
rp = 4K - IlH-d.2 4K ♦
Z
W .d .2 J j where:K = median of the chi-squared distribution with degrees of freedom equal to the number of factors measured (sixteen in the 16 P F )
UJj = Weight or loading on personality factor j. given in Table 7*4
dj = difference in factor j between the mean profiles of the occupations. Based on this formula the following rp values were computed in the
comparison of the profiles of the five occupations each with the profile of OR scientists. (i)
(i) OR Scientists Versus Accountants (England and Scotland) r = -0-055 (NS)1
P