Você está na página 1de 4

loumal of Applied Psychology 1900, \ol SO, Xo 6, 563-566

MOTIVATOR AND HYGIENE DIMENSIONS FOR RESEARCH AND DEVELOPMENT ENGINEERS 1


GEORGE B GRAEX University of Minnesota Herzberg's 2-factor theor\ appears to ofier promising leads to new research on work motivation One of the mam problem-, in following these leads is that the measurement of the work lactors must be accomplished through interviewers The purpose of this stud> was to develop ps\chometnc measure.-, of these work factors through the method ol factor analysis A questionnaire was developed based upon Herzberg'i classification scheme Engineers served as Ss The results show that the dimensions proposed b\ Herzberg when represented as items and rated b> Ss do not result in homogeneous groupings in the fjctoianalvtic sense.

Herzberg, Mausner, and Snyderman (1959) explored job "factors" contributing to satisfaction and dissatisfaction for engineers and accountants. They used the critical incidents method to interview the subjects about previously satisfying and dissatisfying job situations and submitted the interview protocols to a content analysis As a result. Company Policies and Practices, Supervision, Human Relations, and Working Conditions were mentioned more frequently in the stories about bad times than in stories about good times These variables which were related to the work environment were called job-context variables or "hygiene factors." In contrast. Achievement, Recognition, Advancement, Responsibility, and Work Itself were reported more frequently in stories about good times. These variables which were related more closely to the work itself were referred to as job-content variables or "motivators." In general these findings have been replicated by studies employing the story-telling approach (Herzberg, 1965: Schwartz, Jenusaitis, & Stark, 1963). Herzberg and his associates attribute these
1 The computer time for this study was granted b\ the Numerical Analysis Center, University of Minnesota Aid and advice from Rene V. Dawis, Marvin D Dunnette, George W England, and Thomas A Mahone\ are gratefully acknowledged; cooperation of the management and engineers of the two companies is deeply appreciated This paper is based upon a master's thesis submitted to the University of Minnesota, 1963 Part of this paper wa read at the Midwestern Psychological Association, Chicago, April 1965

findings to the fact that job factors realh arc two-dimensional On one dimension, the jobcontent variables when present and favorable tend to result in satisfaction, but their absence does not result in dissatisfaction. On the other dimension, the job-context variables when present and unfavorable tend to produce dissatisfaction, but their absence does not produce satisfaction. According to this two-factor theory, increases in favorable job-context variables will increase job attitudes to a neutral level but only increases in favorable jobcontent variables will increase satisfaction above this level If job variables really do operate in this two-dimensional way, the results yield obvious and important implications for organizational practices In situations where it is desirable to increase positive feelings toward the job, Herzberg's results seem to prescribe that the most efficient means to accomplish this end is to use available resources to improve the job-context variables only enough to remove the irritants from the job environment, and more importantly, to improve the job-content variables as much as possible so as to achieve a high level of job satisfaction. To facilitate such actions, it would be nice to have something other than interviewers' judgments to measure the dimensions, the "motivators" and "hygiene factors " The main difficulty in using Herzberg's categorization procedure to measure job dimensions is that the coding is not completely determined by the rating system and the data, but requires, in addition, interpretation by the rater. For

563

564

GEORGE B. GRAEN

curately as possible the job categories emplo\ed by Hcrzberg when he defined his dimensions (Hcrzberg et al, 1959, pp 143-146) A major determinant of the factors resulting from a factor analysis is the sampling of items Therefore, no attempt was made to improve upon the content of Herzberg's sample, and even attempt was made to represent it In all, 96 items were written to cover the content of Herz-

example, the dimension of Supervision-tech- berg's 16 dimensions Each item was written specifically to measure an attitude corresponding to a nical includes among others the categories: particular job category' An example of how tht (a) "Supervisor competent," (b) "Supervisor questionnaire items were derived is as follows The incompetent," and ( a "Supervisor showed dimension "Status" was defined by al! interview data favoritism " The three categories all call for included in the categories (a) "Signs or appurtenance1 an evaluation of the supervisor's behavior If of status," (b) "Having a given status, ' and <<; "Not having a given status." The respective que>the respondent offers the evaluation, no in- tionnaire items for "Status" were (a) "I receive ,i terpretation by the rater is required. How- symbol of status on my job," (6) "I have a given ever, if the respondent merely describes the status because of my job," and (c) "I do not hau supervisor's behavior, an interpretation by a given status because of my j o b " Both positiu and negative items were developed in order to elicit the rater is necessary. response for both positive and negative feelings toThe necessity for evaluations of the data ward the job by a ratei may lead to contamination of the The item-- required, as did the categories, furthedimensions so derived. Employing a story definitions before they could be rated With content presented by Herzberg to illustrate the di- anaMi* this task was performed by a judge In thistudy the subjects performed the task mension of Recognition, Vroom (165) The 96 items were placed in a format utilizing an pointed out the way in which the two- importance scale ranging from "Very Important to factor theory may contaminate the coding M\ Job Satisfaction" through "Not Important" to procedure. The dimensions in a situation such "Very Important to My Job Dissatisfaction " Below ah this may reflect more the rater's hypotheses is an example of the scale concerning the compositions and interrelations Very Important Not Ver> Important ImporTo My Job To M> Job of dimensions than the respondent's own perSatisfaction tant Dissatisfaction ceptions. Thus, one could conceivably learn 3 4 5 5 4 3 2 1 2 more about the perceptions of raters than Respondents were asked to rate the importance of those of the respondents each job situation to their overall job satisfaction or In view of the possible sources of error in dissatisfaction The 96 items were scored on the dethis categorization procedure, a more objec- gree of importance to overall feelings toward the tive approach would be to have the respon- job Thus the numerical values shown on the scale dents do the rating and perform the necessary were used as scores Respondents were 153 professional engineers workevaluations The respondent's ratings then inc in design and development tor two electronic* could be factor analyzed to identify the fac- firms located in the Twin Cities (Minneapolis-St tors used in the evaluations With this pro- Paul) area The 96 items were intercorrelated, and the 96 X 96 cedure one could be more confident the factors matrix was factor analyzed using the actually represent the perceptual job domain correlation Principal Components method with the Kaiser Variof the respondents rather than the rater's max orthogonal rotation Squared multiple correlainterpretation of the job domain. tions were employed a^ estimates of the comIn short, what is needed to measure the munalities The Kaiser criterion was used to determine the number of factors to extract (Harman, dimensions postulated by Herzberg and his 1960) associates are objective psychometric measRESULTS ures. This study was designed to develop such measures by using the method of factor analyThere were 21 factors, accounting for 61% sis. of the total variance, which were extracted. However, only 11 of these included three or PROCEDURF more items, and these were the ones chosen Questionnaire item* were written to reflect as ac- for further interpretation.'1 Table 1 shows the
-The CD 1604 program was written by Lawrence Liddiard, Numerical Analysis Center, University of Minnesota 3 A table giving the complete correlation matrix has been deposited with the American Documentation Institute Order Document No. 9078 from ADI Auxiliary Publications Project, Photoduphcation Serv-

MOTIVATOR AND HYGIENE DIMENSIONS

565

TABLE 1
DEFINED VARIMAX FACTOR STEDCTTRE FOR ENGINEERS

Factor

FL

Ir

SMC
.12 84 87 .88 .89 .83 .83 .81 83 .10 .84 .83 .83 79 79 72 .70 .07 .87 85 .74 81 .06

Salary- Advancement I receive a wage increase 78 .70 The amount of salary I receive for my job is not adequate 72 78 I expected a wage increase, but did not receive it .72 .74 My wages compare unfavorably with others doing similar or the 71 75 work I expected advancement but I failed to receive it .62 i 75 I receive advancement on my job 59 ! 68 The amount of salary 1 received for my job is adequate 58 ' M\ wages compare favorably with others doing similar or the same work ! 57 ' 70 I am demoted from one position to a lesser position 52 .61 Interpersonal Relationssubordinates and peers I am on good working terms with rm subordinates 78 '' 74 I am on poor personal terms with my subordinates .7d ' ' .73 I am on good personal terms w ith my subordinates 74 .69 I do not like the people I work with on m\ job .65 ', 68 I am on poor working terms w ith my subordinates 68 .65 I I like the people 1 work with on my job There is a lack of cooperation on the part of my co-worken> 44 I 60 Working Conditionsph> sical I have good facilities to work with on my job -.72 76 The physical surroundings of my work place are good - 72 .72 I have poor facilities to work with on my job -.67 62 The physical surroundings of my w ork place are poor 65 - 61 Recognition-Supervision My supervisor shows favoritism to his "special" subordinates . 6 1 65 My work is successful but I arn criticized and punished for n 58 69 My supervisor is unwilling to even listen to my suggestion 55 t ) 5 One of my co-workers gets credit for m\ successful work 51 60 48 My work is successful but I am criticized for it 6 1 Recognition-Achievement My idea is formally accepted by the company -.72 .66 1 proved to my critics that I was right all along - 70 57 My work is praised and special reward is given to me -.43 .56 Work Itself My job gives me an opportunity to perform all phases of an operation 68 | .64 M\ w ork represents a creative or challenging opportunity for me 55 | 55 .54 ' .58 My work is \aned in nature .41 ; .63 My work is routine in nature i Job Secunt\ - 6'/ j 62 I haven't am objective signs that mj job is secure -64 64 I receive objective signs that ny job is secure - 46 j .66 I am in full agreement with the goals of the compain 1 work in Position Itself -.63 ! 65 My work is too easy for me - . 5 8 ! .45 1 have too much w ork to do on m\ job -.50 .63 I receive objective evidence that I am advancing in skill on m\ job Achievement .60 I can not see the results of my own work 62 .51 54 I can see the results of my own work 47 56 1 have a good idea w hich solves a problem Status -.74 64 I ha\ e a given status because of m\ job - 67 64 I do not have a given status because of my job -.56 70 I receive a s) mbol of status on my job Working Conditionsco-workers 7\ 62 My work place is isolated from other people 62 57 I am isolated from the work group on my job . 4 0 . 6 6 I work with and around co-workers on m> job

. 7 5 . 8 2
. 8 0

78 . 7 9
.05
75 .76

.05 78 .75 75 80 .05


75 .79 .79
.05

.83 .67 .78


.04

. //

. 7 5 . 7 3
04
.74 .74

.80
.03

72 . 6 7 77

S'ote.A = 153. FL = Factor loading of statement h> = Calculated (.ommunahty of statement. SMC Squared multiple correlation used as estimate of communahty. p = Proportion of common variance accounted for by factor.

566

GEORGE B. GRAEN

11 factors, the defining items, their factor loadings, communalities, and the proportion of common variance accounted for by each factor.* The factors were given descriptive names based on the names of Herzberg's dimensions which provided the nucleus of defining items. For example, Salary-Advancement was defined by six items from the dimension of Salary and three from that of Advancement. Of the 11 factors, only Job Security and Status included all the items which were originally written to measure them. The next 2 factors in terms of similarity to Herzberg's dimensions were Work Itself and Achievement which included, respectively, four out of six and three out of seven items written to represent them. Only the above 4 factors contained items written to measure one dimension The other factors all included items based on several different dimensions. For example, SalaryAdvancement contained items from the dimensions for Salary and Advancement; Interpersonal Relationssubordinates and peers contained items from Interpersonal Relations subordinates and Interpersonal Relations peers. Also, it is clear that items carefully written to represent specific dimensions end up in quite different factors. Thus, items for Recognition appeared on two separate factors: Recognition-Achievement and RecognitionSupervision. Items for Working Conditions appeared in both Working Conditionsphysical and Working Conditionsco-workers.
DISCUSSION

not constitute homogeneous groupings of job content in the factor-analytic or correlational sense; certainly, the engineers participatinn in this study failed to perceive them as such. The finding that items from a single dimension ended up in different factors and that items from different dimensions ended up in the same factor points up the difficulty inherent in any subjective effort to form categories or "factors" from interview data. Although the use of an a priori classification scheme often appears valid, such use is subject to rater (or judge) bias and error. It ma\ be useful during the exploratory phases of a study to proceed on the basis of an a priori classification scheme, but such a scheme must be treated as a set of hypothesized dimensions. These hypothesized dimensions must be checked out by more objective methods when they are to be employed in studies designed to investigate lawful relationships. In the present study many of the items derived from Herzberg's categories appear not to belong together They did not demonstrate sufficient homogeneity to yield factors. Thi? finding confirms once again the importance of empirical validation before establishing categories as if they were distinct and measurable entities It is suggested that, before the implications of Herzberg's two-factor theory are acted upon, the theory be tested further in carefully designed research utilizing objective measures of the postulated dimensions REFERENCES
HARMAN, H. H Modern factor analysis Chicago University of Chicago Press, 1960 HERZBERG, F The motivation to work among Finnish supervisors Personnel Psychology, 1965, lg, 393402
HERZBERC, F., MAUSNER, B., & SXYDERMAN, B The

It is clear from the foregoing results that the content categories established by Herzberg when represented as items and rated by engineers do not result in factors. They do
ice, Library of Congress, Washington. D C. 20540 Remit in advance SI 75 for microfilm or $2 50 for photocopies and make checki payable to: Chief. Photoduplication Service. Library of Congress 4 A table giving the complete factor structure has been deposited with the American Documentation Institute Order Document No 9078 from ADI Auxiliary Publications Project, Photoduplication Service, Library of Congress, Washington, D C 2OS4O Remit in advance $1 75 for microfilm or S2 50 for photocopies and make checks pa>able to Chief, Photoduplication Service, Library of Congress

motivation to work. (2nd ed ) New York: Wiley. 1959


ScnwARTz, M M, JENUSAITIS, E., & STARK, H

Motivational factors among supervisors in the utility industry Personnel Psychology, 1963, 16, 45-53 VROQM, V H Some observations regarding Herzberg's two-factor theory. Paper read at the American Ps\ chological Association, Chicago, September 1965. (Received February 21. 1966)

Você também pode gostar