Journal of Occupational Psychology, 1979, 52, 129-148. Printed in Great Britain
Scales for the measurement of some work attitudes and aspects of psychological well-being PETER WARR, JOHN COOK and TOBY WALL MRC Social and Applied Psychology Unit, University of Sheffield Two studies of male manual workers are described, in which eight scales relevant to the quality of working life are introduced and assessed. The scales build upon previous work, but are designed to remedy certain conceptual and operational deficiencies. They cover work involvement, intrinsic job motivation, higher order need strength, perceived intrinsic job characteristics, job satisfaction, life satisfaction, happiness, and self-rated anxiety. In addition, components of job satisfaction and life satisfaction, derived through cluster analyses, are also identified. The scales are shown to have good internal reliability and to be factorially separate. Comprehensive psychometric data are provided as a base-line for future applications.
Adequate measurement of complex psychological states usually requires an iterative process; researchers must move several times between conceptualization and operationalization, adjusting their ideas and measures as they go. This is not always feasible within the span of a single research project, and it is sometimes necessary to accept or adapt a previously reported measure or to create a new scale with only limited opportunity for systematic development. Neither approach is entirely satisfactory, especially as most investigators have understandably given priority in their reports to substantive research questions rather than to the provision of detailed information about samples, means, variances, intercorrelations and other features which would assist in subsequent assessment of their measures. This problem is particularly evident in studies of the quality of working life and occupational well-being. The need to examine a large number of subjective variables has often led investigators to devise their own items or to select from previous measures small segments with unknown psychometric properties. An additional difficulty arises from the complexity and ill-defined scope of many concepts in the area; questionnaire items are sometimes difiicult to comprehend, especially for blue-collar workers who are typically the focus of research. There is thus a need for development work to create robust instruments in the quality of working hfe area. Of particular value would be short scales which are easily completed by unsophisticated respondents, which are known to be psychometrically acceptable, and for which normative data are available. This paper contributes towards meeting this need by presenting eight separate scales for diagnostic and evaluative use in both research and practice. These measure work involvement, 129 0305-8107/79/0602-0129502.00/0 © 1979 The British Psychological Society
130
p. W A R R , J. COOK A N D T. W A L L
intrinsic job motivation, higher order need strength, perceived intrinsic job characteristics, job satisfaction, life satisfaction, happiness and self-rated anxiety. The concepts in question have all been important in previous research, and earlier studies will be described next. However, one terminological feature should first be noted. We have used 'job' to refer to the tasks undertaken in a particular setting, whereas 'work' is taken to cover jobs more generally. In this way, for example, 'job motivation' refers to a person's motivation in his or her current post, whereas 'work involvement' deals with involvement in jobs in general. This distinction between job and work has not always been drawn in the literature, with some consequential confusion. WORK INVOLVEMENT AND INTRINSIC JOB MOTIVATION Work involvement may be viewed as a component of the 'protestant work ethic' (e.g. Blood, 1969; WoUack et al., 1971). It emerged as a measurable concept from the paper by Lodahl & Kejner (1965), who presented both a 20-item scale (with 'at least three dimensions') and a more homogeneous version of six items. Subsequent research has indicated statistically significant positive relationships between their measures and age (e.g. Jones et al., 1975), reported participation in decision-making (e.g. Siegel & Ruh, 1973), internality on Rotter's (1966) locus of control scale (Runyon, 1973), endorsement of the protestant ethic (Saal, 1978), and aspects of job satisfaction, especially satisfaction with intrinsic features of a job (e.g. Weissenberg & Gruenfeld, 1968). Gechman & Wiener (1975) have reported a significant correlation with amount of voluntary unpaid overtime; and Warr & Lovatt (1977), using a simple two-item scale, found a significant association with speed of obtaining a new job after redundancy. Involvement is usually found to be unrelated to job performance (see the review by Rabinowitz & Hall, 1977), although there is a suggestion from Pelz & Andrews' (1976) results that among scientists involvement and performance are significantly interrelated (rs mostly between 0-20 and 0-30 in their study of nine samples with n ranging between 65 and 401). This pattern is consistent with other findings that attitudes and job performance are more likely to be associated among managerial and professional employees, who have greater personal opportunity to shape their own job activities than have many manual workers (Warr, 1978c). However, there are still reasons for dissatisfaction with the concept and measures as employed so far (e.g. Lawler & Hall, 1970). The focus of the Lodahl & Kejner items is mainly upon a person's present job (indeed they used the term 'job' involvement), whereas their definition was more in terms of work in general. Second, this definition was itself uncertain. On the one hand, the authors were interested in 'the intemalization of values about the goodness of work or the importance of work in the worth of the person' (p. 24). We view this as the extent to which a person wants to be engaged in work, and take that as our basic definition of work involvement. Lodahl & Kejner's second definition was 'the degree to which a person's work performance affects his self-esteem' (p. 25). This is rather different, being concerned principally with a person's present job and the extent to which he wants to perform well in that job. We prefer to treat this separate notion as intrinsic Job motivation. The term ' intrinsic' is used to emphasize that the motivation is towards personal achievement and task success rather than towards 'extrinsic' satisfactions arising from features
MEASUREMENT OF SOME WORK ATTITUDES
131
such as additional pay or good working conditions. Illustrative items from Lodahl & Kejner's six-item scale are: 'most things in life are more important than work' (a negatively scored statement of work involvement), and 'I'm really a perfectionist about my j o b ' (representing what is here termed 'intrinsic job motivation'). This latter concept is akin to Lawler's (1969) 'intrinsic motivation', defined as 'the degree to which a job holder is motivated to perform well because of some subjective rewards or feelings that he expects to receive or experience as a result of performing well'. Lawler & Hall (1970) tapped intrinsic motivation through items such as ' I feel a great sense of personal satisfaction when I do my job well', and statements of this kind make up the 'internal work motivation' scale of Hackman & Oldham (1975, 1976). Scores on that scale have been found to be significantly positively associated with aspects of job satisfaction and certain perceived job characteristics such as responsibility and knowledge of results (e.g. Hackman & Oldham, 1975; Oldham et al., 1976; Wall et al, 1978). The last-named authors also report a statistically significant relationship (/•=0-41) with employee mental health, measured by the General Health Questionnaire (Goldberg, 1972).
HIGHER ORDER NEED STRENGTH AND PERCEIVED INTRINSIC JOB CHARACTERISTICS Another feature which has attracted recent attention is employees' need for satisfaction and achievement through skilled and autonomous work. The intensity of this need has been variously labelled 'growth need strength' (Hackman & Oldham, 1975), 'self-actualization need strength' (Sims & Szilagyi, 1976) and 'higher order need strength' (Hackman & Lawler, 1971). We will use the last of these terms, recognizing its derivation from Maslow's (1970) hierarchical theory. Higher order need strength has some conceptual similarity with intrinsic job motivation, but as defined here the latter refers only to a specific job situation whereas higher order need strength is viewed as a dispositional characteristic extending across jobs. Measurement of this characteristic has typically been through the items provided by Hackman & Oldham (1975). These include 'would like' ratings of, for example, ' opportunities for personal growth and development in my job' and forced-choice preference responses between items like 'a job for which the pay is good' and 'a job with considerable opportunity to be creative and innovative'. We have found that the content and complexity of these items present difficulties and that a more comprehensible scale is particularly required for blue-collar respondents. Another component of Hackman & Oldham's general model is the extent to which certain 'work motivating factors' are involved in a job. These are frequently measured in terms of ratings of the presence of job variety, autonomy, task identity, task significance and feedback to the worker (see also Hackman & Lawler, 1971), and positive associations between their presence and overall job satisfaction have been summarized by Wall (1978). However, the five factors often turn out to be moderately intercorrelated (median interfactor values of 0-52, 0-20 and 0-42 are reported by Dunham, 1976, Steers & Spencer, 1977 and Saal, 1978, respectively), and it would be useful to obtain standardization data for a brief single measure of the presence of job characteristics which might give rise to intrinsic satisfaction. We will refer to this as a scale oi perceived intrinsic job characteristics.
p . WARR, J. COOK AND T. ^
Studies ofjob satisfaction have taken many different forms since the original work of Hoppock (1935). Some investigators have preferred to obtain overall satisfaction scores, either through single items (e.g. Quinn et al., 1974) or through aggregates of several items (e.g. Brayfield «fe Rothe, 1951). Other researchers have used instruments specially constructed to tap a number of different features of satisfaction. For example, the Job Description Index (Smith et al., 1969) contains subscales to measure attitudes towards pay, promotion prospects, the work itself, supervision, and co-workers. The Worker Opinion Survey (Cross, 1973) has in addition a subscale to tap feelings about the firm as a whole, and a similar instrument for use with managers also covers attitudes towards subordinates (Warr & Routledge, 1969). These measures have a number of disadvantages. They tend to contain redundant and overlapping items and are rather long and cumbersome. Their emphasis has been primarily upon extrinsic features of satisfaction, to the relative exclusion of intrinsic components. Furthermore, they sometimes confuse descriptive and evaluative judgements, although it is only the latter which can be said genuinely to measure satisfaction (e.g. Payne ef a/., 1976). LIFE SATISFACTION, HAPPINESS AND SELF-RATED ANXIETY Psychological well-being is a diffuse concept which deserves greater measurement attention (e.g. Bradburn, 1969; Warr & Wall, 1975). One set of investigations has examined the nature and correlates of life satisfaction, usually construing this in terms of people's expressed satisfaction with features of their environment and everyday life. Studies from a 'social indicators' perspective have examined national survey samples' ratings of a wide range of features (your car, your house, your sleep, your freedom of speech, etc.) (e.g. Andrews & Withey, 1974, 1976; Hall, 1976a). Some research has given emphasis to the interrelationships between life satisfaction and other features of psychological well-being and mental health (e.g. Bradburn, 1969; Warr, 1978i); and other studies have looked at both life satisfaction and job satisfaction. For example. Hall (1976ft) reported an intercorrelation of 0-42 in a British sample; and in an American study London et al. (1977) observed a median correlation of 0-21 between single job satisfaction items and an overall life satisfaction measure. These latter authors also observed that job satisfaction contributed more to life satisfaction for men than for women, a finding which echoes Hulin's (1969) observation. A measure which has often been employed in national surveys (e.g. Quinn & Shepard, 1974) is an expression of personal happiness on a three-point scale. This provides another perspective on well-being and in view of its simplicity and the availability of comparison data is often worth recording. Research into anxiety has traditionally been directed towards measures of trait anxiety, neuroticism, etc., and there is a need for short indices of self-rated anxiety. A number of possibilities have been examined by Bradburn (1969), Warr (1978ft) and others, but (as with life satisfaction) further investigations with particularly salient items are desirable. SUMMARY OF TERMS
Working definitions of the several concepts outlined above are as follows. The numbers assigned to each are employed consistently throughout the paper.
MEASUREMENT OF SOME WORK ATTITUDES
133
1. Work involvement is viewed as the degree to which a person wants to be engaged in work. 2. Intrinsic job motivation is viewed as the degree to which a person wants to work well in his or her job in order to achieve intrinsic satisfaction. 3. Higher order need strength is taken to be the importance which a person attaches to the attainment of higher order needs. 4. Perceived intrinsic Job characteristics are the person's reports about the degree to which features are present in his or her job which might give rise to intrinsic satisfaction. 5. Job satisfaction is the degree to which a person reports satisfaction with intrinsic and extrinsic features of the job. Total job satisfaction is the sum of all separate items, and overall job satisfaction is reported satisfaction with the job as a whole. 6. Life satisfaction is the degree to which a person reports satisfaction with salient features of his life and life-space. Total life satisfaction is the sum of all separate items, and overall life satisfaction is reported satisfaction with one's life as a whole. 7. Happiness is the degree to which a person reports that he or she is currently happy. 8. Self-rated anxiety is the degree to which a person reports anxiety about salient features of his or her life and life-space summed across items, and overall self-rated anxiety is reported anxiety in general. Definitions of components of concepts 5 and 6 will be introduced later.
Scales to measure these eight concepts were developed through two interview studies with blue-collar workers. The initial pool of items was drawn from the literature and through discussion, and a pilot study was previously carried out to assist with initial decisions about items and procedures. Study 1 took place in February 1977 and Study 2 in November 1977. The Samples The respondents in the two principal studies were 200 and 390 blue-collar male workers within the mainland United Kingdom. They were all in full-time employment (thus excluding the self-employed), had worked in their present job for at least a month, and had a mean length of service of 9-02 years. They were aged between 20 and 64, and were all employed in manufacturing industry (orders 3 to 19 of the Standard Industrial Classification, which excludes service industries, construction, agriculture, mining and transport). Respondents were drawn in equal numbers from 10 widely dispersed sampling areas (Study 1) and 20 areas (Study 2) according to a predetermined frame which approximately matched national demographic characteristics. This specified that half the sample in each area should be from firms employing fewer than 300 employees and half from larger companies (see, for example. Department of Employment, 1978). Within each half of the sample, 50 per cent should be above and 50 per cent below 40 years of age, and within each quarter-sample half should be skilled, 30 per cent semi-skilled and 20 per cent unskilled. Decisions about skill level were made in terms of training required before a person was judged competent at his job (months or years, several weeks, or a few days respectively for the three levels of skill), and examples of jobs in each category were supplied to interviewers from the material provided by the Office of Population Censuses and Surveys (1970).
134
p. WARR, J. COOK AND T. WALL
Other biographical information was gathered during the interview, and the two samples were of almost identical composition. A combined analysis (« = 590) revealed that 13 per cent were unmarried, 24 per cent were married without children at home and 63 per cent were married with children at home. In terms of length of service with their present employer, 26 per cent reported less than 2 years, 22 per cent between 2 and 5 years, 24 per cent between 5 and 10 years, 17 per cent between 10 and 20 years, and 11 per cent more than 20 years. Nine per cent reported that their company contained no trade union members; and 'some', 'most' and 'all' company employees were reported to be trade unionists by 12, 33 and 46 per cent of respondents respectively. Procedure Interviews were concerned solely with the scales under investigation, and were carried out individually within respondents' homes by trained female staff of National Opinion Polls Ltd. Each interviewer sought volunteers to complete her quota sampling frame. The average time for each interview was about 30 minutes; in Study 2 this included other scales which are not described here. Interviewers indicated that respondents had 'very little' or ' n o ' difificulty with the material and that they generally appeared to enjoy the interview. Interviewers read out the instructions and items to each person, who selected his answer from the set of alternatives listed on a card. The interviewer then recorded the response in numerical form on the questionnaire. The Items The research design required decisions to be taken on the basis of Study 1 results so that shorter and better scales could be used in Study 2. The latter study, with a larger number of respondents, was intended to provide cross-validation evidence and to yield reliable norms for this population. In order to include additional scales in Study 2 (not reported here) the Life Satisfaction scale was omitted from this second investigation. The items finally selected and the response dimensions employed are presented in full in Appendix A; the same sequence of scales was used in both studies. It will be seen that seven-point responses were sought throughout, except for section 4 {perceived intrinsic job characteristics) where five alternative responses were employed and for the happiness item which involved a three-point response scale. Scoring was from 1 to 7, 1 to 5 or 1 to 3 throughout, with 1 being the most negative response in each case. Each scale or subscale score was the unweighted sum of the responses to the included items.
Decisions about exclusion of items in Study 1 were based upon inter-item and item-whole correlations (desired to be high within a scale), mean scores (desired to be away from the end-point), standard deviations (desired to be high), and the meaning of each item (excessive redundancy within a scale was undesirable). The final number of items in each scale is shown in Table 1. These were achieved after omission of 2, 2 and 4 items in scales 1 to 3 respectively. Scale 4 in Study 1 comprised seven items, but five additional ones were included for Study 2 in order to achieve
MEASUREMENT OF SOME WORK ATTITUDES
til SlfS
?5SSSS
S?J5 I I
III
i>|S
n
iillllillillllllillill r:csicO^
LOlSSlSS
lg.2cD.S2
^£^C6^
136
p . WARR, J. COOK AND T. WALL
more extensive coverage of intrinsic job characteristics; two of these were subsequently omitted in the light of the Study 2 results, leaving a 10-item scale. The final scales contain no negatively keyed items, although some were included in the initial battery. This group of items scaled only marginally less reliably than did the positively keyed ones, but they were excluded from Study 2 because they were reported by interviewers to present conceptual difficulties for some respondents. In addition to the eight scales already identified, it will be seen that the left-hand column of Table 1 contains several separately numbered and labelled items. Three of these (5x, 6x and 8x) are the single-item overall job satisfaction, life satisfaction and anxiety reports sohcited at the end of scales 5, 6 and 8 (see Appendix A). The other terms newly introduced in Table 1 are subscales of measures 5 and 6, derived through cluster analyses using the furthest neighbour method. The job satisfaction items (scale 5) are divided in two separate ways. At one level of analysis, two separate clusters of items (5a and 5b) could readily be identified. Seven items (numbers 5.2, 5.4, 5.6, 5.8, 5.10, 5.12, 5.14 in Appendix A) came together into a subscale which is appropriately termed intrinsic job satisfaction (5a). The other items represent extrinsic job satisfaction (5b). However, the full set of items could also be viewed at a different level in terms of three other clusters (5c, 5£/and 5e). One group of four items (5.2, 5.6, 5.8, 5.14) emerged as a cluster specially concerned -with job itself intrinsic satisfaction (labelled 5c). Another cluster of five extrinsic satisfaction items (5.1, 5.3, 5.5, 5.13, 5.15) appeared to be best described as working conditions extrinsic satisfaction (5d). The remaining items (5.4, 5.7, 5.9, 5.10, 5.11, 5.12) comprised a cluster which straddled the intrinsic and extrinsic features in a way which suggested a concern for individual recognition and management behaviour; we have interpreted this in terms of employee relations satisfaction (5e). Cluster analysis of the life satisfaction items yielded an interpretable threecomponent structure. Items 6.5, 6.6, 6.9 and 6.10 of scale 6 in Appendix A deal with immediate personal concerns about health, education, social and family life, and we have referred to this cluster as satisfaction with personal life (6a). Seven other items (6.7, 6.8, 6.11, 6.12, 6.13, 6.14, 6.15) form a cluster to do with satisfaction with standards and achievement (6b); and the third cluster of four items (6.1, 6.2, 6.3, 6.4) covers satisfaction with life style (6c). The choice of the complete scale or subscales in any investigation will depend upon the degree of specificity which is required. The subscales are strongly intercorrelated (see Table 3), and they are of course statistically associated with the full scales of which they are part. A number of features in Table 1 deserve comment. The values presented for Study 1 are derived from the final set of items, after exclusion of less satisfactory material. They are therefore directly comparable with the Study 2 values, except for scale 4 where additional items were present in Study 2; results for the incomplete scale 4 used in Study 1 are therefore omitted from the table. In all cases the scales' internal homogeneity is good, as represented by alpha coefficients (Nunnally, 1967) and mean item-whole correlations, and these values are closely replicated across the two studies. Furthermore, the individual item-whole correlations (not shown in the table) remain very similar across the studies. The rank-order correlations between item-whole values for each item in a scale on the two occasions average around 0 95. The standard deviations remain very similar in the two studies, with the exception of scale 3 (higher order need strength), which showed a significant decrease (P< 0-001)
MEASUREMENT OF SOME WORK ATTITUDES
137
from 6-80 to 5-03. This may have occurred because of the omission of four items from the original Study 1 version of this scale, generating a more homogeneous set of statements in Study 2. (Recall that the values in Table 1 cover the same items for each study.) There are small but statistically significant increases in mean scores for the first three scales from Study 1 to Study 2. These may also have arisen partly from the omission of the more divergent items; as noted above omissions were only made from these three scales. The decline in scores for scale 8 (self-rated anxiety) is not statistically significant, but is of interest in that examination of individual items revealed a large decrease in only one case, anxiety about Britain's economic future. This refiects a marked change in the economic climate between the dates of the two studies, illustrated in sample newspaper headlines like ' Food prices shock and worse to come', 'Few will escape higher bills' (February, 1977) and 'Just the job—boost from the dole queues' and 'Taxes to be cut next week' (November, 1977). It is also notable that the mean scores are located away from the end-point of the scales. It is often observed, for example, that the large majority of employees report themselves satisfied with their job (e.g. Quinn et al., 1974). Mean job satisfaction scores in the present studies are above the scale mid-point, but closer to the mid-point than to the maximum. This is also the case for life satisfaction, with the exception of subscale 6b where the mean value is below the scale mid-point; on the other hand, scales 1, 2 and 3 have relatively high means. The pattern of correlations within and between the scales was closely replicated across the two studies. This may be illustrated by the factor analysis results in Table 2. These show the varimax-rotated loadings of each item on the six scales common to both studies. (Note that the 'overall' items identified as Sx, 6x, 1 and 8x in Table 1 were not included.) In keeping with an assumed model of six independent constructs, the first six factors only were rotated in each study, these embracing 50-1 and 46-5 per cent of the accounted variance in the two cases. It can be seen that scale 1 {work involvement) straddles the first two factors in the Study 1 results, but that otherwise the factor structure is remarkably consistent with the model. Items in each scale all load highly on a separate factor, and their loadings are extremely similar in the two studies. A summary of the interrelationships between the total scale and subscale scores is presented in Table 3. This matrix derives from the combined sample of 590 respondents, except for the values associated with scale 4 {perceived intrinsic job characteristics) and scale 6 {life satisfaction). Since the final version of scale 4 contains more items than were used in Study 1, it is more appropriate to cite the correlations with the final 10-item scale observed in Study 2 (« = 390). On the other hand, as scale 6 was not used in Study 2, the values associated with this measure in Table 3 are based upon the Study 1 sample of 200 respondents. It can be seen that age is uncorrelated with other variables in the table except scale 2 {intrinsic job motivation) where r=O-ll. This contrasts with the small but significant relationships between similar variables and age which have been reported elsewhere (e.g. Saleh & Otis, 1964; Aldag & Brief, 1977; Rabinowitz & Hall, 1977; Saal, 1978). As would be expected, skill level is moderately associated (r=0-29) with scale 4 {perceived intrinsic job characteristics), and to a lesser extent with higher order need strength (r = 0-17). Correlations were also calculated with length of service, family responsibility and degree of unionization. These are not shown in Table 3, but
p . WARR, J. COOK AND T. WALL
MEASUREMENT OF SOME WORK ATTITUDES
^,-OOCMr-
525ll
o CO in <- i i O CM CO •* I I
255;;;§
in CM p^ co r^ m m co CM co o CM r-. O O .-O O O O T- ,-O »-O »-O O I I I I I I I f I I
r. t^ c CN CM O
ooooo" III I
SJ2!:3 l o S S o I I ocM^-cM^cMO^cNOococoT-cDi I I
p . WARR, J. COOK AND T. WALL
CNiT-r^050ocococDO>or^^
I
c^g
.«!
Illill l|f«l 11
MEASUREMENT OF SOME WORK ATTITUDES
141
no Statistically significant relationships with scale scores were observed. Expected intercorrelations between age and length of service (r=0-44) and organization size and degree of unionization (r=0-27) were however present. Scales 1 and 2 (work involvement and intrinsic job motivation) represent the two facets of Lodahl & Kejner's (1965) 'work involvement'. In the case of the present measures they are moderately intercorrelated (r=0-37), and relate in a similar fashion to the other variables. Note however that the latter scale is more strongly associated with higher order need strength and perceived intrinsic job characteristics (0-33 and 0-36 against 0-17 and 0-28). Scale 3 (higher order need strength) is itself in addition significantly associated only with scales 4, 6a and 8. Scale 4 (perceived intrinsic job characteristics) reveals expected strong associations with job satisfaction (especially intrinsic satisfaction) and smaller but significant relationships with life satisfaction and happiness. It should be noted that the life satisfaction scale used in this research contained no items directly relevant to work. The job satisfaction clusters are closely related to each other, suggesting that separate subscale scores will not always warrant calculation. The significant relationships with life satisfaction (e.g. 0-42 between the two total scores) and with happiness and self-rated anxiety (0-49 and -0-24 respectively) are however of interest. The subscales of life satisfaction are less strongly related to each other than are those of job satisfaction, and this aspect of well-being is, as expected, significantly correlated with happiness but is only just significantly associated with self-rated anxiety. Different aspects of psychological well-being are being tapped by the three scales (see also Warr, 1978ft). Appendix B presents the means, standard deviations and decile scores for each scale and subscale except scale 7 where the three-point response dimension makes decile scores inappropriate. Results for this scale (happiness) have often been presented in terms of the percentage of respondents giving each of the three answers. Thus an American national sample (n=2145) yielded 37-4 per cent 'very happy', 56-4 per cent 'pretty happy', and 6-2 per cent 'not too happy' (Quinn & Shepard, 1974). Observed values for this sample of British male manual workers in manufacturing industry were 27-7, 60-7 and 11-6 per cent respectively. (Note that we used 'fairly happy' instead of 'pretty happy'.) A final question concerns the test-retest reliability of the measures. Sixty participants in Study 2 were re-interviewed 6 months later. These were drawn from among those who had remained in their previous job to constitute a sample which matched the earlier group in terms of region, age and skill level. Test-retest correlation coefficients were as follows: work involvement 0-56; intrinsic job motivation 0-65; higher order need strength 0-26; perceived intrinsic job characteristics 0-69; total job satisfaction 0-63; overall job satisfaction 0-27; self-rated anxiety 0-63; overall self-rated anxiety 0 49. The satisfaction subscales yielded values of 0-68, 0-51, 0-60, 0-45 and 0-64 for 5a to 5e respectively. The observed test-retest reliability of the higher order need strength scale (0-26) is undesirably low. This concept has presented problems of operationalization to other investigators (cf. Steers & Spencer, 1977) and there is clearly scope for further improvement. The low stability of the single overall job satisfaction item (0-27) may be contrasted with the high value of 0-63 for the 15-item scale of total job satisfaction. All the other coefficients are acceptably high, relative to measures in the literature and to the internal homogeneity of the scales. It should also be noted that the mean values
p . WARR, J. COOK AND T. WAL
observed on the two occasions did not differ significantly from each other, nor did the standard deviations except for scales 1 and 3 which were slightly smaller (P<005)
DISCUSSION The eight measures described here have proved to be acceptable to blue-collar workers and their psychometric properties appear to be good. The main purpose of this paper has been to describe the measures and to provide base-line data for subsequent applications, but construct validity evidence in terms of expectations from the literature would of course be helpful. The intercorrelations between scales, with one or two exceptions, are similar to those reported in North American research with other measures. Previous studies, summarized in the Introduction, have suggested that work involvement may be more closely associated with intrinsic than extrinsic satisfaction, whereas the difference was not observed here (/•=0-27 and 0-30 respectively). One explanation arises from the alternation of intrinsic and extrinsic items in our job satisfaction measure and the resulting strong intercorrelation between the two subscales (r=0-72). However, the correlations between work involvement and fob satisfaction of between 0-23 and 0-30 are typical of those in the literature (e.g. McKelvey & Sekaran, 1977; Rabinowitz & Hall, 1977; Saal, 1978). PForA: invo/ve/Me«/is usually found to be less strongly related to higher order need strength (e.g. Rabinowitz & Hall, 1977; Saal, 1978), and the observed value in Table 2 is 0-17. Intrinsic Job motivation is also related to Job satisfaction (median r=0-33) and perceived intrinsic Job characteristics (r = 0-36), in line with evidence presented in the Introduction. The latter measure is strongly related (0-73) to total Job satisfaction and to intrinsic Job satisfaction (0-76), as expected. Total Job satisfaction is moderately associated with total life satisfaction (0-42) and happiness (0-49) and less strongly but significantly associated with self-rated anxiety (-0-24). It should be emphasized again that in order to ensure independence from other measures the life satisfaction scale used here did not make any reference to work; one or more additional items to cover this feature may be appropriate in some other studies of life satisfaction. Another extension of the present method would be to use the scales in a self-completion mode; as described in the Method section, responses were here recorded by interviewers. The scales are currently being employed in several other investigations and are proving to be extremely appropriate for self-completion; indeed, item simplicity was one objective in their construction. One illustrative self-completion study (conducted in collaboration with Robin Maurice) examined mailed responses from 340 university graduates some 6 months after leaving university. Their mean job satisfaction scores were higher than for the blue-collar workers studied here: 74-61, 34-32 and 40-37 for scales 5, 5a and 5b (SD 11-39, 6-53 and 6-23 respectively); but the values for men (« = 200) and women («= 140) were extremely similar. Intrinsic and extrinsic Job satisfaction were found to intercorrelate 0-65 for the full sample, and total Job satisfaction was associated with satisfaction with personal life and satisfaction with life-style at the 0-36 and 0-28 level respectively. Subscale 6b was not used in full, but the summed response to items 6.7 and 6.8 in Appendix A (about accomplishments and the future) correlated 0-61 with total Job satisfaction. A total life satisfaction score from items 6.1 to 6.10 was found to
MEASUREMENT OF SOME WORK ATTITUDES
143
correlate 0-49 with total job satisfaction. One interesting feature was that the associations between job and life satisfactions for the men in this sample were consistently higher than those for women; however, the differences (e.g. 0-53 and 0-39 for total life satisfaction) did not attain statistical significance. Finally, a general comment about scale homogeneity is in order. In all these studies the items in each scale were presented in a block before moving on to the next scale (see Appendix A). This is of course a widespread practice and on balance we favour it over the randomization of items from different scales. It is however likely to encourage internal homogeneity and may reduce associations between scales; these outcomes should be borne in mind when instruments are being compared.
REFERENCES AxDAG, R. J. & BRIEF, A. P. (1977). Age, work values and employee reactions. Industrial Gerontology, 4, 192-197. ANDREWS, F . M . & WITHEY, S . B . (1974). Developing measures of perceived life quality: Results from several national surveys. Social Indicators Research, 1, 1-26. ANDREWS, F . M . & WITHEY, S . B . (1976). Social Indicators of Well-being. New York: Plenum Press. BLOOD, M . R . (1969). Work values and job satisfaction. Journal of Applied Psychology, 53, 456-459. BRADBURN, N . M . (1969). The Structure of Psychological Well-being. Chicago: Aldine. BRAVFIELD, A . H . & RoTHE, H. F . (1951). An index of job satisfaction. Journal of Applied Psychology, 35,307-311. CROSS, D . (1973). The Worker Opinion Survey: A measure of shop-floor satisfactions. Occupational Psychology, 47, 193-208. DEPARTMENT OF EMPLOYMENT (1978). How big is British business? DE Gazette, 86, 37-40. DUNHAM, R . B . (1976). The measurement and dimensionality of job characteristics. Journal of Applied Psychology, 61,404-409. GECHMAN, A . S. & WIENER, Y . (1975). Job involvement and satisfaction as related t o mental health and personal time devoted to work. Journal of Applied Psychology, 60, 521-523. GOLDBERG, D . P. (1972). The Detection of Psychiatric Illness by Questionnaire. Oxford: Oxford University Press. HACKMAN, J. R. & LAWLER, E . E . (1971). Employee reactions t o job characteristics. Journal of Applied Psychology, 55, 259-286. HACKMAN, J. R. & OLDHAM, G . R . (1975). Development of the Job Diagnostic Survey. Journal of Applied Psychology, 60,159-170. HACKMAN, J. R. & OLDHAM, G . R . (1976). Motivation through the design of work: Test of a theory. Organizational Behavior and Human Performance, 16, 250-279. HALL, J. (1976a). Subjective measures of quality of life in Britain: 1971 to 1975. Social Trends, 7, 47-60. HALL, J. (19766). The quality of life. Unpublished research report. HoppocK, R. (1935). Job Satisfaction. New York: Harper. HuLiN, C. L. (1969). Sources of variation in job and life satisfaction: The role of community and job-related factors. Journal of Applied Psychology, 53, 279-291. JONES, A. P., JAMES, L . R . & BRUNI, J. R. (1975). Perceived leadership behavior and employee confidence in the leader as moderated by job involvement. Journal of Applied Psychology, 60,146-149. LAWLER, E . E . (1969). Job design and employee motivation. Personnel Psychology, 22, 426-435. LAWLER, E . E . & HALL, D . T . (1970). Relationship of job characteristics to job involvement, satisfaction, and intrinsic motivation. Journal of Applied Psychology, 54, 305-312. LoDAHL, T. M. & KEJNER, M . (1965). The definition and measurement of job involvement. Journal of Applied Psychology, 49, 24-33. LONDON, M . , CRANDALL, R . & SEALS, G . W . (1977). Contributions of job and leisure satisfaction to quality of life. Journal of Applied Psychology, 62, 328-334. MASLOW, A . H . (1970). Motivation and Personality, rev. ed. New York: Harper & Row. MCKELVEY, B . & SEKARAN, U . (1977). Toward a career-based theory of job involvement: A study of scientists and engineers. Administrative Science Quarterly, 22, 281-305. NuNALLY, J. (1967). Psychometric Theory. New York: McGraw-HUl.
144
p. WARR, J. COOK AND T. WALL
OFFICE OF POPULATION CENSUSES AND SURVEYS (1970). Classification of Occupations. London: HMSO. OLDHAM, G . R . , HACKMAN, J. R. & PEARCE, J. L. (1976). Conditions under which employees respond positively to enriched work. Journal of Applied Psychology, 61, 3 9 5 ^ 0 3 . PAYNE, R . L . , FINEMAN, S. & WALL, T . D . (1976). Organizational climate and job satisfaction: A conceptual synthesis. Organizational Behavior and Human Performance, 16, 45-62. PELZ, D . C . & ANDREWS, F . M . (1976). Scientists in Organizations, rev. ed. Michigan: Institute for Social Research. QuiNN, R. P. & SHEPARD, L . J . (1974). The 1972-73 Quality of Employment Survey. Michigan: Institute for Social Research. QuiNN, R. P., STAINES, G . L . & MCCULLOUGH, M . R . (1974). Job Satisfaction: Is there a Trend? Washington: US Department of Labor. RABINOWITZ, S . & HALL, D . T . (1977). Organizational research on job involvement. Psychological Bulletin, 84, 265-288. ROTTER, J. B. (1966). Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs, 80, whole no. 609. RuNYON, K. E. (1973). Some interactions between personality variables and management styles. Journal of Applied Psychology, 57, 288-294. SAAL, F . E . (1978). Job involvement: A multivariate approach. Journal of Applied Psychology, 63, 53-61. SALEH, S. D . & OTIS, J. L. (1964). Age and level of job satisfaction. Personnel Psychology, 1 7 , 4 2 5 ^ 3 0 . SiEOEL, A. L. & RuH, R. A. (1973). Job involvement, participation in decision making, personal background and job behavior. Organizational Behavior and Human Performance, 9, 318-327. SIMS, H . P. & SZILAGYI, A. E. (1976). Job characteristic relationships: Individual and structural moderators. Organizational Behavior and Human Performance, 17, 211-230. SMITH, P. C , KENDALL, L . M . & HULIN, C . L . (1969). The Measurement of Satisfaction in Work and Retirement. Chicago: Rand McNally. STEERS, R . M . & SPENCER, D . G . (1977). The role of achievement motivation in job design. Journal of Applied Psychology, 62, 472-479. WALL, T . D . (1978). Job redesign and employee participation. In P. B. Warr (ed.). Psychology at Work, 2nd ed. Harmondsworth, Middx: Penguin. WALL, T . D . , CLEGG, C . W . & JACKSON, P. R. (1978). An evaluation of the job characteristics model. Journal of Occupational Psychology, 51, 183-196. WARR, P. B. (1978a). Attitudes, actions and motives. In P. B. Warr (ed.). Psychology at Work, 2nd ed. Harmondsworth, Middx: Penguin. WARR, P. B. (19786). A study of psychological well-being. British Journal of Psychology, 69,111-121. WARR, P. B. & LOVATT, D . J. (1977). Retraining and other factors associated with job finding after redundancy. Journal of Occupational Psychology, 50, 67-84. WARR, P. B. & ROUTLEDGE, T . (1969). An opinion scale for the study of managers' job satisfaction. Occupational Psychology, 43, 95-109. WARR, P. B. & WALL, T . D . (1975). Work and Well-being. Harmondsworth, Middx: Penguin. WEISSENBERG, P . & GRUENFELD, L . W . (1968). Relationship between job satisfaction and job involvement. Journal of Applied Psychology, 52, 469-473. WOLLACK, S., GooDALE, J. G., WiJTiNG, J. P. & SMITH, P. C. (1971). Development of the survey of work values. Journal of Applied Psychology, 55, 331-338. Received 10 July 1978 M R C Social and Applied Psychology Unit Department of Psychology University of Sheffield Sheffield SIO 2TN
MEASUREMENT OF SOME WORK ATTITUDES
145
APPENDIX A The instructions and finally-selected items are presented in full below. The sequence shown is that employed in the research, but the numbering has been altered to correspond with the numbers used in the text. The scales may be freely used, and minor modifications to the instructions may sometimes be required. However, users are requested to supply to the authors as much detail as possible about samples, results and correlates.
Scale 1: Work Involvement Introduction. For some people work is just a means to get money, it's something they have to put up with. For others, work is the centre of their life, something that really matters to them. I would first of all like to ask you about your reactions to work in general, and whether actually doing work is important to you personally. By 'work' I mean having a paid job. Here are some statements which people have made about work and working, in general. Without limiting yourself to your present job would you indicate on this scale (SHOW CARD 'W') how strongly you agree or disagree with each comment in turn? Remember that I'm asking about paid jobs in general, not simply your present job. 1.1. Even if I won a great deal of money on the pools I would continue to work somewhere 1.2. Having a job is very important to me 1.3. I should hate to be on the dole 1.4. I would soon get very bored if I had no work to do 1.5. The most important things that happen to me involve work 1.6. If unemployment benefit was really high I would still prefer to work Scale 2: Intrinsic Job Motivation Introduction. Now can we move in a little closer to how you personally feel about your present job? Again I would like you to think about a number of statements that people have made about work, but this time think about your present job, not work in general. Please indicate on the same scale as before (SHOW CARD 'W') how strongly you agree or disagree with each comment. Remember that I'm asking now about your present job as a (INSERT TITLE). 2.1. I feel a sense of personal satisfaction when I do this job well 2.2. My opinion of myself goes down when I do this job badly 2.3. I take pride in doing my job as well as I can 2.4. I feel unhappy when my work is not up to my usual standard 2.5. I like to look back on the day's work with a sense of a job well done 2.6. I try to think of ways of doing my job effectively.
Scale 5: Job Satisfaction Introduction. The next set of items deals with various aspects of your job. I would like you to tell me how satisfied or dissatisfied you feel with each of these features of your present job. Each item names some aspect of your present job as a (INSERT TITLE). Just indicate how satisfied or dissatisfied you are with it by using this scale (SHOW CARD 'X'). 5.1. The physical work conditions 5.2. The freedom to choose your own method of working 5.3. Your fellow workers 5.4. The recognition you get for good work 5.5. Your immediate boss 5.6. The amount of responsibility you are given 5.7. Your rate of pay 5.8. Your opportunity to use your abilities 5.9. Industrial relations between management and workers in your firm 5.10. Your chance of promotion
146
P. WARR, J. COOK AND T. WALL
5.11. 5.12. 5.13. 5.14. 5.15. 5x.
The way your firm is managed The attention paid to suggestions you make Your hours of work The amount ofvariety in your job Your job security Now, taking everything into consideration, how do you feel about your job as a whole?
Scale 4: Perceived Intrinsic Job Characteristics Introduction. You may have felt in the last section that some of the job features mentioned were not present in your job very much. It is likely that some of the aspects did apply to your job, while others applied less or not at all. Could we now go through a small number of these items again, together with a few new ones, but this time thinking about how much you feel each feature is present in the job you are doing? For this we use a different scale (SHOW CARD 'M'). 4.1. The freedom to choose your own method of working 4.2. The amount of responsibility you are given 4.3. The recognition you get for good work 4.4. Being able to judge your work performance, right away, when actually doing the job 4.5. Your opportunity to use your abilities 4.6. The amount of variety in your job 4.7. Your chance of promotion 4.8. The attention paid to suggestions you make 4.9. The feeling of doing something which is not trivial, but really worthwhile 4.10. Doing a whole and complete piece of work Scale 3: Higher Order Need Strength Introduction. Now let's look at the things that matter to you in a job. What things are important in a job and what things are less important in your opinion ? I'd like you to think about paid work in general—any paid job you might do or might like to do, not just your present job. I'm going to mention a number of characteristics which you might look for in a job. Please show me on this scale (SHOW CARD ' Y') how important each one is when you think about jobs you would like to have. 3.1. Using your skills to the maximum 3.2. Achieving something that you personally value 3.3. The opportunity to make your own decisions 3.4. The opportunity to learn new things 3.5. Challenging work 3.6. Extending your range of abilities Scale 8: Self-rated Anxiety Introduction. So far we have thought a great deal about work and your job. For this set of items I would like you to consider some wider aspects of your life that go beyond work, although they may include it. Most people these days have something to worry about, sometimes big things, sometimes quite small things. Would you think back over the past few weeks and let me know to what extent you may have been concerned or worried about various circumstances that affect your life. This is the scale to be used for this section (SHOW CARD 'Z'). 8.1. Not having enough money for day to day living 8.2. Your immediate family 8.3. Your health 8.4. Growing old 8.5. How things are going at work 8.6. Britain's economic future 8x. In general, how worried or concerned do you feel these days ? Scale 6: Life Satisfaction Introduction. Finally, will you consider some other aspects of your life at the present moment.
MEASUREMENT OF SOME WORK ATTITUDES and indicate tisfied you feel about each one ii ? Please use this scale again (SHOW CARD 'X'). 6.1. The house or flat that you live in 6.2. The local district that you live in 6.3. Your standard of living: the things you can buy and do 6.4. The way you spend your leisure time 6.5. Your present state of health 6.6. The education you have received 6.7. What you are accomplishing in life 6.8. What the future seems to hold for you 6.9. Your social life 6.10. Your family life 6.11. The present government 6.12. Freedom and democracy in Britain today 6.13. The state of law and order in Britain today 6.14. The moral standards and values in Britain today 6.15. Britain's reputation in the world today 6x. Taking everything together, your life as a whole these days.
Scale 7: Happiness And, as a final item. 7. Taking all things together, how v auld you say things were these days ? Would you say yoi 3. Very happy 2. Fairly happy 1. Not too happy
Card' fV' (for Scales 1 an( 2) 1. No, I strongly disagree 2. No, I disagree quite a lot 3. No,Idisagreejustalittle 4. I'm not sure about this 5. Yes, I agree just a little 6. Yes, I agree quite a lot 7. Yes, I strongly agree
Card'X' (for Scales 5 and 6) 1. I'm extremely dissatisfied 2. I'm very dissatisfied 3. I'm moderately dissatisfied 4. I'm not sure 5. I'm moderately satisfied 6. I'm very satisfied 7. I'm extremely satisfied
Card' Y' (for Scale 3) 1. Not at all important 2. Not particularly important 3. I'm not sure about its import! 4. Moderately important 5. Fairly important 6. Very important 7. Extremely important
Carrf'Z'(for Scale 8) 1. Not at all concerned 2. Just a little concerned 3. Mildly concerned 4. Worry a little 5. Quite worried 6. Very worried 7. Extremely worried
Card'M'(for Scale 4) 1. There's none of that in my job 2. There's/«.s/ a little of that in my job 3. There's a moderate amount of that in my job 4. There's quite a lot of that in my job 5. There's a great deal of that in my job.
WARR, J. COOK AND T. WALL APPENDIX B Deciles, means and standard deviations for all see is and subscales for the combined sample (/7=590) (except for scale 4 (/7=390) and scale 5 ( A ) = 2 0 0 ) ; see text). Respondents were blue-collar male employees in anufacturing industry.
Decile
1 2 3 4 5 6 7 8 9 10
SD
1
2
Work involvement
Intrinsic job
240 ZIZ 299 31 8 333 350 362 373 39-1 420
285 323 342 356 370 38-1 394 40 8 41-4 420
594
3
strength
4 Perceived intrinsic job characteristics
266 31 0 330 347 358 370 384 398 41 3 420
208 245 275 298 322 350 373 40-1 43 5 51 0
476 568 636 68 0 71 5 756 785 835 890 1020
203 252 28-1 31 4 332 349 372 392 425 490
5-80
839
1542
8-25
Higher order
551
5
Intrinsic job
faction
faction
Working Job itself conditions Employee Overall intrinsic extrinsic relations job satisfaction satisfaction satisfaction satisfaction 130 159 177 19 2 205 21 8 230 239 257 280 2032 490
6b Satisfaction with andards and Decile 2 3 \ 5 3 7 3 3 10 Mean SD
147 183 21 5 242 256 272 288 308 337 42 0 2544 7-12
182123242527282931352589 484
6c Satisfaction with life style 148 169 180 19-1 20-1 21 2 22-1 230 239 28 0 2022 386
5a
Total job
258 304 336 362 382 40 3 424 447 48 0 560 836
Total life satisfaction 533 57-1 620 644 670 704
12-4 16 7 200 230 247 267 285 306 342 420 2440 795
56 Extrinsic job satisfaction
19 8 207 21 5 222 229 239
930 6709 11 40
6x Overa life satlsfact 61 Ml 132 '•153 ^175 I194 27 64 i 04 ^00 •13 •28
90 138 157 174 191 21 4 242 278 420 1861 7-18
100 125 1 69 2^14 262 3^10 358 412 511 700 337 V59