Attitude Dimensions of Leisure: A Replication Study

Abstract

A previous paper reported the construction of instrument for the measurement of leisure attitudes and the resulting identification of five relatively independent dimensions in the leisure domain (Neulinger and Breit 1969). In line with an elemental principle of competent (Kessen 1960) and the fact that these findings were based on a factor-analytic procedure, often unstable phenomenon, it was thought that a replication study was in order before any further use of the above instrument was made. The purpose of the present paper is to report the results of this replication study. In addition, since the results showed the dimensions to be extremely stable and thus useful for future research purposes, available covariates of the identified dimensions will be presented.MethodSubjectsThe sample consisted of 335 adults working full-time, 198 males and 137 females, ranging in age from eighteen to sixty-eight, with a mean age of thirty-five. Thirty-six percent of the respondents were Jewish, 26 percent Catholic, 15 percent Protestant. Eighteen percent reported no religious preference, Respondents were predominantly white (90 percent), the majority married, and their educational level quite high (median category some college), although all levels were represented. Reported average family income was equally high: in the $11,000 to $13,000 bracket. Eighty-nine percent of the respondents were born in the United States. The spectrum of occupations was very broad and included the professions, business, industry and the trades.QuestionnaireThe questionnaire consisted of eighty attitude items plus questions relating to standard socio-economic variables. Thirty-two of the items were included for the purpose of replication; twenty-seven of these were identical to the highest loading items in each of the previously identified five leisure dimensions. Five of the items were new and had been written to get at the essence of these factors. The balance of the items related to the leisure environment, the conceptualization of leisure and work, and to sexual attitudes.ProcedureData were collected in the Spring of 1970, in and around New York City. Respondents were obtained through the cooperation of The City College students. Great emphasis was placed on appealing to the respondents' cooperation and honesty, and they were assured complete anonymity.ResultsA factor analysis of thirty-two items, using a principal components method and Varimax rotation, yielded five factors. The five factor solution was in line with theoretical expectations, and also justified in terms of the criterion that the number of eigenvalues greater than one determines the number of factors (Harman 1960). Five factors accounted for 38 percent of the total variance. An inspection of the results and comparison with the previous analysis reveals amazing degree of similarity of outcomes (Table 1). The same labels have been used to identify the dimensions: (1) affinity for leisure, (2) society's role in leisure planning, (3) self-definition through leisure or work, (4) amount of perceived leisure, and (5) amount of work or vacation desired. As before, only those items were used to define factors that had loadings of at least .30, and .10 larger on the respective factor than on any other factor. Of the five new items, two loaded on the factors they had been written for, namely Factors III and IV. The item written for Factor V turned up on Factor I, loading to a lesser degree on Factor V. The item written for Factor I tied on Factor V and was thus excluded. The item written for Factor II did not relate to any of the factors.Only two of the previous twenty-seven items load now on a different factor than before. They are the two lowest loading items, now in Factor I (Doing nothing, being idle, ... ; taking of habit forming drugs), previously in Factor II. Overall, then, it is quite apparent that the dimensions identified are stable across different samples. …

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 44 citations exist upstream.