Beruflich Dokumente
Kultur Dokumente
Commitment Strength*
R o n a l d C. W i m b e r l e y
North Carolina State University
This paper suggests that strength of commitment be measured apart from the underlying
commitment positions. This suggestion is empirically directed to the study of religious commit-
ments where much conceptual and empirical effort has been spent. Guttman attitudinal compo-
*Comments and suggestions by Louis E. Dotson are appreciated. Data were made available by the
Education Division and the Research and Statistics Department of the Southern Baptist Convention.
211
212 SOCIOLOGICAL ANALYSIS
In the belief dimension, Glock and Stark (1968: 57) propose to categorize individuals by the degree of
commitment to religious beliefs. However, in their measure of the ritual dimension fora nationwide
sample, Stark and Glock ( 1968:104-105) combine questions on the general importarlce of religion with
an item on commitment to a religious behavior. They also combine frequency of private prayer with
importance ofprayer to measure private religious ¡ (Stark and Glock, 1968:121). Likewise, Lenski
(1961:57) conceptually associates orthodox belief with being strongly committed to religious beliefs.
He also defines devotionalism as the importance of personal communication with God, yet
operationalizes the concept as types of religious behavior positions (Lenski, 1961:25,57).
2This is certainly not offered as a rigorous example ofa Guttman scale since there were only two items.
However, assurance of scalability was slightly improved by trichotomizing the frequency of the item on
prayer. The item on asking help in everyday matters was only dichotomized because of a preponder-
ance of responses in two categories. Their coefficient of reproducibility (Edwards, 1956:184-188) is
.91. Not only does this coefficient meet the conventional requirement of a .90 level, but it compares
favorably to the minimum marginal reproducibility (Edwards, 1956:191-193) of .64 for these items
and response catego¡ Likewise, the Borgatta (1955) error ratio for this scale, .56, is quite low as
desired. Therefore, the items on devofionalism are not found to be unscalable and their scaling is
thought to representan improvement of Lenski's procedure of summated scoring.
MEASUREMENT OF COMMITMENT STRENGTH 213
3The .96 coefficient of reproducibility (Edwards, 1956:184-188) is well above the .72 minimum
marginal reproducibility (Edwards, 1956:191-193). The Borgatta (1955) error ratio of .52 is also
considered quite satisfactory. It is noted that in terms of Guttman components, this construction of a
scalable intensity measure is apparenfly unique. Suchman (1950) was unable to establish ah intensity
scale and there is no known evidence of success since his try.
4King (1967:176) and King and Hunt (1979) note that the religious experience dimension which they
produccd from factor and cluster analyses was similar to Glock's (Glock and Stark, 1965) experiential
dimension and Lenski's (1961) ctevotionalism orientation. Since then, Stark and Glock (1968: 15,
108-124) have used devotionalism asa different conceptual dimension of religious commitment from
expe¡ Whereas they regard devotionalism as an element of the ritualistic dimension which
designates private, informal dimension of religious ritual behavior including prayer and Bible read-
ing, the experience dimension pertains to one's communication with adivine being. This conceptual
distinction seems to draw a very fine line. In fact, the product moment correlations between measures
of these two dimensions were the fourth highest for Protestants, +.467, and the second highest for
Catholics, +.424, in 36 such correlations among religious dimension measurements for each group
(Stark and Glock, 1968:177-178). Nudelman ( 1971) reports a factor analysis of these scale intercorrela-
tions and finds expe¡ and devotionalism in the same general dimension for the Protestant and
Catholic samples. Although a factoring of scales is notan adequate test of the dimensionality of their
items, Nudelman's factoring of items from Christian Scientist data produce the same results.
To the extent that the devotionalism items on prayer and consultation with God involve communica-
tive experience with adivine being, these items seem equally appropriate as indicators of private
religious ritual or experience commitments. At any tate, the conceptual distinction is moot for
purposes at hand. The items used here on frequency of communication with God and items on
importance of this communication are the operational substances, regardless of whether such com-
munication is called devotionalism or experience.
214 SOCIOLOGICALANAI.YSIS
FIGURE 1
Content and Intensity
Medians
Intensity Percentiles
High
62 7
J
//
J
51
Intensity Percentiles
40
Midpoint
Content Percentiles
23' 43 78
Low Content Percentiles High
TABLE 1
havior rather than attitudes, the linear relationship might be due to the behavioral
content. Had there been enough persons in the sample displaying infrequent
communication behaviors, which would have warranted several scale types in
place of the single lowest position shown for these data, the line of relationship
might have extendedlower and further to the left. In other words, a greater range
of content positions probably would have revealed weaker commitments a n d a
lower intercept.
MEASUREMENT OF COMMITMENT STRENGTH 215
Conclusions
For a better understanding of religious commitment and its strength, it would
be beneficial to know if this relationship is similar for different scales of religious
commitment or religious commitment strengths, for different samples, and over a
period of time.
Studies by Bahr, Bartel, and Chadwick ( 1971) and by Gibbs, Mueller, and Wood
(1973) have already found religious salience and orthodoxy to be useful predic-
tors. However, in each case an inspection of their salience measures finds many
items relative to things other than orthodoxy. Ir is proposed here that when
specific strength measures are linked to corresponding commitment dimensions,
predictions from religious factors should improve markedly.