Early career job satisfaction for full-time workers who are deaf or hard of hearing.

This study presents the results of a quantitative study of job satisfaction (i.e., the degree to which an employee feels good or bad about his or her job) for a sample of deaf and hard of hearing alumni of postsecondary education programs. Hearing-impaired people comprise almost ten percent of the nation's working-age population (ages 16-64). Within this population, about 500,000 individuals became deaf before age 19, 2.2 million became deaf between age 20 and 64, and about 9.6 million people aged 16-64 are hard of hearing. These figures were derived by the authors based on prevalence estimates provided by the National Center for Health Statistics (Reis, 1994) and estimates of the size of specific age groups within the general population (U.S. Bureau of the Census, 1991).

A comprehensive literature search involving electronic searches and contacts with VR and postsecondary education experts netted 13 published and unpublished empirical studies of job satisfaction for people who are deaf or hard of hearing. Generally, these 13 studies report that most study participants indicate they are satisfied or very satisfied with their jobs, with relatively few participants reporting being dissatisfied (Bigman, 1960; Boatner, Stuckless, & Moores, 1964; Christiansen, 1982; El-Khiami, 1993; Grant, Marron, & Welsh, 1981; Grant & Welsh, 1981; Johnson, 1993; Justman & Moskowitz, 1963: Lunde & Bigman, 1959; Mowry, 1986; Schein, 1968; Schein & Delk, 1978; Vescovi, 1973). This pattern has been observed repeatedly even under such diverse research methodologies as mail surveys and interviews conducted in sign language.

Five of the studies reported results from subgroup comparisons. Individuals who are highly satisfied with their employment tend to have more hearing ability and are more orally competent (Bigman, 1960). Among graduates of the National Technical Institute for the Deaf (NTID), those with a bachelor's degree were generally more satisfied with their jobs than those with postsecondary certificates (Grant, et al., 1981). Mowry (1986) interviewed 22 deaf former vocational rehabilitation clients. He found that those who earned more money and with less job tenure were more satisfied. Grant and Welsh (1981) concluded from a study of deaf NTID alumni that females tended to be more satisfied than males with various aspects of a job, such as satisfaction with pay, co-workers, and the job in general. El-Khiami (1993) surveyed 490 deaf or hard-of-hearing alumni from 47 special postsecondary programs. She found that the least satisfied workers were machine operators followed by technical or administrative support (clerical) workers; the most satisfied were workers in professional or managerial occupations.

Each of the aforementioned studies described participant's levels of job satisfaction or the relative levels of job satisfaction for subgroups. This focus stems largely from the specific reasons for which postsecondary institutions and VR agencies study job satisfaction. In both settings, measures of job satisfaction are included among a broader array of outcome measures used to assess the effectiveness of delivered services. Nevertheless, both postsecondary institutions and VR agencies can benefit even more from the additional ability to predict the levels of satisfaction for their consumers as well as from an improved understanding of the effects job satisfaction might have on other work-related outcomes. Such information will not be available Until researchers begin to report the results of studies designed to identify variables which, at minimum, are correlated with measures of job satisfaction, and preferably, are either predictors of job satisfaction or are outcomes of job satisfaction. That is, rather than descriptive studies, multivariate studies are needed.

This paper presents a constructive reanalysis of data from one of the aforementioned studies (El-Khiami, 1993) using multivariate procedures to analyze data. Specifically, this study explores the magnitude and direction (positive or negative) of the relations between job satisfaction and age, gender, hearing status, unemployment history, socioeconomic quality of the job (SEI), total 1993 income, on-the-job work limitations, total number of jobs held, and availability of a teletypewriter telephone device (TTD). The literature on the relation between these variables and job satisfaction is inconclusive or absent for people who are deaf or hard of hearing as well as for the general population. As such, no a priori expectations are offered concerning these relations. The study also included one additional variable which was viewed as dependent upon job satisfaction--job search behavior. Studies (e.g., Mobley, 1982) have shown that employees with lower levels of job satisfaction are known to be more likely to be searching for another job than are employees with higher levels of job satisfaction. Therefore, it is predicted that job satisfaction will be negatively related to job search behavior. No a priori expectations are offered with respect to other relations with job search behavior.

Method

Subjects

The sample consisted of 322 full-time employees who were deaf (69%) or hard of hearing (31%). Subjects were drawn from a sample of 495 deaf and hard-of-hearing alumni of 47 special postsecondary programs, including two- and four-year colleges and vocational-technical institutes, who participated in a longitudinal study of their work experiences. Participants were members of the graduating classes of 1983, 1984, and 1985. Complete work history information was collected five years after graduation for each respondent.

The primary dependent variable is job satisfaction; therefore, only full-time employees were considered to be eligible for this study. Excluded were subjects who were unemployed or out of the labor market. These criteria eliminated potential memory errors for those unemployed subjects who had responded to questions about previous jobs. Part-time employees were excluded due to their small number (n = 27) and concerns about the comparability in findings between full- and part-time workers. A list-wise method for deleting cases with missing data reduced the sample from 347 to 322 subjects (93% of eligible subjects). This sample is smaller than in the El-Khiami (1993) study which utilized the same data base; thus, comparisons with the results in El-Khiami would be inappropriate.

The 322 subjects ranged in age from 24 to 60, with a mean age of 28.5. Both the relatively young average age of our subjects, and the relatively low divorced/widowed rate (given below), are consistent with our focus on early career job satisfaction. Males (55%) outnumbered females (45%). Eight percent of the subjects were from minority groups. Just over half (56%) of the respondents were never married, 40% were married, and four percent were divorced or widowed. In terms of education, 36% had postsecondary vocational certificates or diplomas, 23% had associate's degrees, 26% had bachelor's degrees, five percent had master's degrees, and the remaining 15% did not report their most recent degree attainments. The average annual income (both mean and median) from all sources (such as job, social security) fell in the "$20,000 to $24,999" category.

Measures

That the typical student who is deaf in grades eight (8) through twelve (12) reads at about a third grade level (Allen, 1986) is a factor influencing questionnaire development and test design. Our college-educated sample reads at an estimated fifth or sixth grade level on average. For this reason, concise and readable questions were developed and pilot tested to measure job satisfaction, job search behavior and other study variables.

Job satisfaction was measured by asking "Overall, how satisfied are you with the job you have now...?" Response options were 1="not satisfied" to 2="satisfied" to 3="very satisfied." The job satisfaction question is similar in wording (simplified wording) to an overall job satisfaction question studied by Scarpello and Campbell (1983) who reported that an item worded "How satisfied are you with your job in general" did not lack reliability and demonstrated a level of predictive validity matching and exceeding other single and multiple-item measures of overall job satisfaction. Job search behavior was measured by asking "Are you looking for a job now?" Response options were 1 = "yes" and 0 = "no." The job search behavior question is a similar, yet simplified, version of a job search behavior item developed by Hom and Griffeth (1991).

The remaining variables included hearing status (1 = "deaf" and 0 ="hard of hearing"), age (chronological age measured in one-year increments), gender (1 = "male" and 0 = "female"), and number of jobs (a count of the total number of jobs held after leaving the postsecondary program including current job, ranging from 1 = "one" to 4 = "four or more" in increments of one). Economic measures include total 1993 income (income from all sources before deductions, measured in 10 categories beginning at "below $9,999" to "50,000 or more" in increments of $5,000), unemployment history (1 = "ever unemployed more than one month after graduation" and 0 = "never unemployed more than one month," and Stevens and Hyon Cho's (1985) Socioeconomic Index (SEI), a measure of the socioeconomic quality of jobs. The observed SKI scores ranged from a low of 13.93 to a high of 85.53. The SKI is based on a composite of educational requirements and job salary (Stevens & Hyun Cho, 1985) and is determined by identifying the SKI scores associated with the job titles from participant's jobs (respondent's job titles were determined using titles provided by participants or by expert judgement based on reading of participant-provided job descriptions). Two other study variables dealt with the work situation: TDD available ("Is a TDD available at work?"; 1 = "yes" and 0 = "no"), and on-the-job limitations ("On the job you have now...is there any part of your job that you could not do because you cannot hear well?"; 1 = "yes" and 0 = "no").

Procedures and Analyses

The surveys were administered in 1988 and 1989 to respondents from the graduating classes of 1983, 1984, and 1985. Response rates for the classes of 1983, 1984, and 1985 ranged from 80% to 85% Data were analyzed using the SPSS-PC package. Means, standard deviations, bivariate correlations and multivariate regression analyses were performed. Two-tailed tests of the significance of the correlation coefficients are reported.

Results

Bivariate Correlations

Table 1 displays the means, standard deviations, and bivariate correlations. The negative relation between job satisfaction and job search behavior was expected (r = -.50; p .001) showing that subjects with low job satisfaction were more likely to be searching for other jobs. Table 1 shows that job satisfaction also covaries to a significant degree with SEI scores (r = .19; p .01), on-the-job limitations (r = -.11; p .05), and unemployment history; r = -.13; p .01). Significant correlations between job search behavior and both TDD available (r = -.13; p .05) and total 1993 income (r = .18; p .001) were also observed.

Table 1
Descriptive Statistics and Intercorrelations (N=322)

Variables                     M       SD     1     2     3     4

1. Age                      28.69    4.34
2. Gender                    1.45     .5     05
3. On-the-job limitations    1.56     .5    -10   -05
4. Hearing Status            1.31     .46   -05    07   -01
5. TDD Available             1.40     .49   -14    05   -02    37
6. Numbers of Jobs           2.29     .49   -14    05    08    10
7. Unemployment History      1.66     .47    05    04   -01   -01
8. Total Income-1993         4.73    1.94    20   -12   -13   -05
9. SEI                      43.05   20.00    28    03    09    06
10. Job Satisfaction         1.79     .68    01    04   -11   -02
11. Search Behavior          1.71     .45    01   -01   -04    02

Variables                    5     6     7     8     9     10

1. Age
2. Gender
3. On-the-job limitations
4. Hearing Status
5. TDD Available
6. Numbers of Jobs           13
7. Unemployment History     -02   -25
8. Total Income-1993        -24   -10    24
9. SEI                       25   -01   -05   -41
10. Job Satisfaction         05   -07   -13   -11    19
11. Search Behavior         -13   -08    08    18   -08   -50

Note: Correlations are reported with decimal points omitted. See under measures for coding of variables. Correlations [is greater than or equal to] .18 are significant at the .001 level. Correlations [is greater than or equal to] .13 are significant at the .01 level. Correlations [is greater than or equal to] .11 are significant at the .05 level. All tests are two-tailed.

Multiple Regression Analyses

Multiple regression analyses were performed to determine the degree to which study variables, taken collectively, would contribute to the prediction of job satisfaction and job search behavior. Job search behavior was regressed onto all other study variables, and job satisfaction was regressed onto all study variables except for job search behavior. Results reported in Table 2 exclude age, gender, hearing status, and number of jobs because these variables did not contribute to the predictions of job satisfaction or job search behavior either collectively or individually. When job satisfaction is the dependent variable, the variables TDD available, unemployment experience, on-the-job limitations, SKI scores, and total 1993 income, taken as a set, account for a modest, yet significant amount of explained job satisfaction variance ([R.sup.2] = .06; F (1,317) = 3.50; p .01); significant contributions are made by the variables of unemployment history, SKI scores, and on-the-job limitations. When job search behavior is the dependent variable, the results in Table 2 show that the set of variables makes a moderate contribution to explained job search behavior variance ([R.sup.2] = .28; F (1,317) = 3.15; p .01), largely due to the unique contribution of job satisfaction, and to a lesser degree, total 1993 income. Earlier a significant relation between TTD available and job search behavior (see Table 1) was noted. However the contribution of TTD available to the prediction of job search behavior is insignificant when considered in the same equation as the other variables, suggesting that further attention to this relation may be unwarranted. The six predictors account for 28 percent of the explained variance in job search behavior.

TABLE 2 Results of Multiple Regressions Predicting Job Satisfaction and Search Behavior (N=322)

                        Standardized Regression Coefficients

Independent Variables   Job Satisfaction   Job Search Behavior

TDD Available                 .01                -.09
Unemployment History         -.11(*)              .03
On-the-Job Limitations       -.14(*)              .06
Total Income - 1993           .02                 .15(**)
SKI                           .18(*)             -.04
Job Satisfaction                                  .50(***)

Summary Statistics

[R.sup.2]                     .065                .28
df                           5,316               5,316
F (Step)                     3.50(**)            3.15(**)

(*) p .05; (**) p ,01; (***) p .001.

Discussion

The observed significant negative correlation between job satisfaction and job search behavior supported the hypothesis that the relation between job satisfaction and job search behavior would be negative in direction. Armed with this knowledge, counselors know that (a) dissatisfied employees are likely to act on their negative feelings, indicating that it is important for counselors to be aware of consumers' feelings about jobs and (b) employed consumers who seek help with a job search may be doing so due to dissatisfaction with their job, rather than due to career aspirations, meaning that the counselor may be in a position to help the client work out the problems at work which stimulated dissatisfaction.

With respect to the possible correlates of job satisfaction, this study indicated that three variables covary with job satisfaction: SKI scores, unemployment history and on-the-job limitations. More research is needed to determine if these findings are reliable; nevertheless, if reliable, such information does have implications for proactive efforts to place consumers in satisfying jobs or to advise students about the merits of certain areas of postsecondary training relative to other areas. The significant relation between SEI scores and job satisfaction shows that consumers should be advised that advanced degrees open doors to better, more satisfying jobs. With respect to the relation between on-the-job limitations and job satisfaction, consumers should be advised of possible on-the-job limitations (attributable to hearing levels) and of strategies for coping with them. For example, consumers may benefit from knowing how to request reasonable accommodations. In turn, the VR counselor may want to educate employers about any special needs a particular consumer may have in performing the essential functions of a job, and then provide some tips for how to accommodate those needs.

A counselor's knowledge of the negative relation between unemployment history and job satisfaction may be useful as well. Counselors should anticipate the negative feelings about working that are likely to be harbored by consumers who have experienced prolonged periods of unemployment. These feelings may need to be addressed prior to initiating a job search, perhaps by building the self-esteem of the consumer who has been unemployed.

With respect to the correlates of job search behavior, beyond job satisfaction, one other study variable was important--total 1993 income. There is a tendency for those with lower total incomes to be searching for another job. It seems that workers who are deaf or hard of hearing make decisions to look for other jobs based on their level of job satisfaction, their overall financial circumstance, and, undoubtedly, other factors not studied here. Thus it is important for counselors to determine why a consumer has decided to search for another job before facilitating the search effort.

Limitations

Readers should note that information about the validity and reliability of the measures of job satisfaction and job search behavior is not available, largely due to the use of single item measures. This does not mean that these measures lack validity or reliability; rather, estimates of validity and reliability simply are not available to present here. However, as noted in the Methods section of this paper, the items were similar in wording to items known to be reliable and valid. Moreover, study findings are consistent with study hypotheses, and therefore the measures have produced the type of predictable findings one would expect from valid measures. Nevertheless, the lack of validity estimates should be considered when interpreting findings.

This study focuses on alumni of postsecondary programs and the results may not be representative of deaf and hard-of-hearing people without a postsecondary education. Further, the study focused on a subset of the potential correlates and consequences of job satisfaction, leaving much room for additional research. This narrow range of correlates may account for the modest observed improvement over guessing at participant's levels of job satisfaction. Nevertheless the results provide new information which may be useful to researchers and practitioners. The multivariate design provides a model for gathering such information.

This article was supported by a grant (No.: H133B10001) from the National Institute on Disability and Rehabilitation Research (NIDRR). The opinions and conclusions expressed are those of the authors and do not necessarily represent policies or endorsement by NIDRR.

References

Allen, T. E. (1986). Patterns of academic achievement among hearing-impaired students: 1975 and 1983. In A. N. Schildroth and M. A. Karchmer (Eds.) Deaf Children in America. San Diego, CA: College-Hill Press.

Bigman, S. K. (1960). Occupations of the deaf. Rehabilitation Record. Nov.-Dec., 23-26.

Boatner, E. B., Stuckless, E. R., & Moores, D. F. (1964). Occupational status of the young adult deaf of New England and the need and demand for a regional technical-vocational training center. (VRA/DHEW RD 1295-S-64). West Hartford, CT: American School for the Deaf.

Christiansen, J. (1982). The socioeconomic status of the deaf population: A review of the literature. In J. B. Christiansen & J. Egelston-Dodd (Eds.), Social aspects of deafness, Volume 4: Socioeconomic status of the deaf population. 161. Washington, DC: Gallaudet College.

El-Khiami, A. (1993). Employment transitions and establishing careers by postsecondary alumni with hearing loss. The Volta Review, 95 (4), 357-366.

Grant, D. J., Marron, M. J., & Welsh, W. A. (1981). Beginning a longitudinal analysis: A look at patterns of labor mobility among deaf graduates of the Rochester Institute of Technology. (Report No. 37). Rochester, NY: Rochester Institute of Technology, National Technical Institute for the Deaf, Institutional Planning and Research.

Grant, D. J., & Welsh, W. A. (1981). Beginning a longitudinal analysis: A look at job success among deaf graduates of the Rochester Institute of Technology over two years. (Report No. 38). Rochester, NY: Rochester Institute of Technology, National Technical Institute for the Deaf, Institutional Planning and Research.

Hom, P.W. & Griffeth, R.W. (1991). Structural equations modeling test of a turnover theory. Journal of Applied Psychology, 76, 350-366.

Johnson, V. (1993). Factors impacting the job retention and advancement of workers who are deaf. Volta Review, 95(4), 341-356.

Justman, J. S., & Moskowitz, S. (1963). A follow-up study of graduates of the school for the deaf. New York: Bureau of Educational Program Research and Statistics.

Lunde, A. S., & Bigman, S. K. (1959). Occupational conditions among the deaf. Washington, D. C.: Gallaudet College.

Mobley, W. (1982). Employee turnover: Causes, consequences, and control. Reading, MA: Addison-Wesley.

Mowry, R. L. (1986). Job satisfaction of deaf former vocational rehabilitation clients. Unpublished manuscript. Little Rock, AR: University of Arkansas, Arkansas Rehabilitation and Training Center on Deafness and Hearing Impairment.

Reis, P. W. (1994). Prevalence and characteristics of persons with hearing trouble: United States, 1990-91. Vital and Health Statistics (Series 10, No. 188). Department of Health and Human Services Publication No. (PHS) 94-1516.

Scarpello, V. & Campbell, J.P. (1983). Job satisfaction: Are all the parts there? Personnel Psychology, 36, 577-600.

Schein, J. D. (1968). The deaf community study of metropolitan Washington, D. C. Washington, D. C.: Gallaudet College.

Schein, J. D., & Delk, M. T. (1978). Economic status of deaf adults. In J. D. Schein (Ed.), Progress Report No.12 (Project No. 16-p-56802/2-12). New York: New York University, Deafness Research and Training Center.

Stevens, G., & Hyun Cho, J. (1985). Socioeconomic indexes and the new 1980 Census Occupational Classification Scheme. Social Science Research. 14, 142-168.

U.S. Bureau of the Census. (1991). Statistical abstract of the US: 1991. (111th ed.). Washington, D.C.: US Government Printing Office.

Vescovi, G. M. (1971). Factors related to successful employment of adult deaf workers. Journal of Rehabilitation of the Deaf. 6(3), 1-11.

Paul D. Geyer, Research Assistant Professor, Rehabilitation Research and Training Center for Persons who are Deaf of Hard of Hearing, Department of Rehabilitation University of Arkansas, 4601 W. Markham St., Little Rock, AR, 72205. E-mail: [e-mail].