'Can You Help Me With Homework? Needed By Friday 4/29/2022

BASIC AND APPLIED SOCIAL PSYCHOLOGY, 1995, 16(1 & 2), 211-225 Copyright © 1995, Lawrence Erlbaum Associates, Inc.

Does the Experience of Organizational Justice Mitigate the Invasion of Privacy Engendered by Random Drug Testing?

An Empirical Investigation

Bennett J. Tepper College of Business and Economics

University of Kentucky

Charles K. Braun College of Business Marshall University

This research used theories of organizational justice to develop predictions regarding employees' judgments of the invasiveness of random-drug-testing programs. An investigation with two firms suggested that employees view random drug testing to be less invasive when they (a) hold management positions, (b) have been tested for drug use fewer times, (c) perceive the consequences of testing positive for drug use to be less punitive, and (d) perceive the drug-testing procedures to be more accurate. Unexpectedly, however, the perceived danger of impaired performance on the job was unrelated to the criterion. Implications of the findings for research and practice are discussed.

The consequences of employees holding negative attitudes toward corporate drug-testing programs include resentment and antagonism (Greenfield, Karren, & Giacobbe, 1989), cheating and refusals to take drug tests (Axel, 1990), and reduced performance (Konovsky & Cropanzano, 1991). Given the proliferation of drug-testing programs and the climate of governmental support for invasive drug-testing procedures (particularly for employees in safety-sensitive jobs; Christopher, 1991), surprisingly little scholarly re- search has explored the attitudes of individuals who undergo random drug testing. Nevertheless, industry surveys suggest that employees and em- ployers generally oppose random drug testing (Axel, 1990), and a recent

Requests for reprints should be sent to Bennett J. Tepper, Department of Management, College of Business and Economics, 355-J B&E Building, University of Kentucky, Lexington, KY 40506-0034.

2 1 2 TEPPER AND BRAUN

Study suggests that most union members view random drug testing to be a serious invasion of privacy (LeRoy, 1991).

However, laboratory and simulation studies suggest that individuals' attitudes toward drug testing depend on contextual characteristics and characteristics of the program (e.g., Crant & Bateman, 1990; Murphy, Thornton, & Prue, 1991; Murphy, Thornton, & Reynolds, 1990; Stone & Kotch, 1989). Employees hold more favorable attitudes toward drug testing when they experience a sense of organizational justice (i.e., the benefits of participation in the drug-testing program outweigh the costs and the drug-testing procedures conform to Leventhal's, 1980, rules of procedural fairness; Crant & Bateman, 1989). Hence, it may be speculated that the experience of organizational justice mitigates the deleterious effects of random drug testing on employee attitudes (cf. Konovsky & Cropanzano, 1991). Given this issue's importance for research and for practice, a field study was undertaken to determine to what extent justice perceptions attenuate the invasiveness of random-drug-testing programs.

THEORETICAL FRAMEWORK: ORGANIZATIONAL JUSTICE

Organizational-justice approaches focus on individuals' reactions to the experience of distributive justice (i.e., the degree to which outcomes are fairly distributed) 2ind procedural justice (i.e., fairness in the processes used to allocate outcomes, and the consistency with which they are applied across individuals; Leventhal, 1980). Whereas individuals experience distributive justice when outcomes are more favorably allocated, procedural justice hinges on the accuracy, consistency, and ethicality of the allocation or decision-making methods themselves (Greenberg, 1990b). Theory and empirical research suggest that the experiences of distributive and proce- dural justice are related to favorable overall attitudes toward a variety of organizational referents including grievance systems (Gordon & Fryxell, 1989), jobs (Alexander & Ruderman, 1987), pay (Folger & Konovsky, 1989), and performance appraisals (Greenberg, 1986). Extending justice theory to the present context, it may be predicted that employees who experience a sense of organizational justice will view random drug testing to be less invasive than will individuals who do not experience a sense of organizational justice; that is, organizational justice should mitigate the invasion of privacy resulting from participation in random drug testing.

ORGANIZATIONAL JUSTICE AND ATTITUDES TOWARD DRUG TESTING

What factors create the experience of organizational justice in the context of drug testing? Crant and Bateman (1989) developed a model that explains

RANDOM DRUG TESTING 2 1 3

how the experiences of distributive and procedural justice affect individuals' attitudes toward drug testing programs. According to the model, distribu- tive justice is enhanced when individuals believe that the benefits of participating in the drug testing program outweigh the costs. Organizational characteristics (e.g., safety, size, interdependency, and performance) and employee characteristics (e.g., demographic variables, norms and expecta- tions, and personality variables) determine the extent to which individuals perceive that drug testing is warranted in a given setting. As examples, employees are more likely to perceive a need for drug testing when:

1. Safety concerns are prominent. 2. Drug abuse is a problem within the organization. 3. Individuals depend on their co-workers to perform efficiently. 4. The organization has a history of high absenteeism and theft, or low

productivity. 5. Individuals belong to a cohort group that accepts the legitimacy of

drug testing.

Crant and Bateman (1989) went on to argue that the experience of procedural justice is a function of the drug-testing program itself; fairness judgments are enhanced when drug-testing procedures conform to Leventhal's (1980) rules of procedural fairness: accuracy, correctability, consistency, bias suppression, and ethicality. As examples, employees are more likely to hold favorable overall attitudes toward drug-testing pro- grams when:

1. The drug-testing procedures accurately discriminate users from nonusers.

2. Individuals who receive treatment or rehabilitation can have their record expunged.

3. Individuals do not feel that they have been singled out for drug testing.

4. The consequences of testing positive for drug use are not excessively punitive.

In the following theoretical analysis, we develop specific predictions about a subset of variables that are likely to affect perceptions of the invasiveness of random-drug-testing programs using Crant and Bateman's (1989) model as a conceptual foundation.

Factors Affecting Distributive Justice

Job characteristics. Of the contextual variables that determine whether or not individuals perceive a need for drug testing, perhaps the

2 1 4 TEPPER AND BRAUN

most important is the danger of impaired performance on the job. Most firms that test for drug use are motivated by safety concerns (Axel, 1990) 2ind the rationale for drug testing should be most apparent to employees in safety-sensitive positions (i.e., the extent to which impaired performance could endanger oneself, one's co-workers, or the public). In two studies of the acceptability of drug testing in 35 different occupations. Murphy et al. (1991) found that individuals held more favorable general attitudes toward drug testing of employees in more safety-sensitive jobs. Hence, although random drug testing usually engenders perceptions of invasiveness, these effects should be mitigated by perceptions that impaired performance on the job could be a hazard to oneself, one's co-workers, or the public. Therefore, it may be hypothesized that employees who perceive their job to be more dangerous will perceive random drug testing to be less invasive than will employees who perceive their job to be less dangerous (Hypothesis I).

Employee characteristics. Employees are more likely to perceive a need for drug testing when participation in the drug-testing program is consistent with accepted ways of thinking and behaving within particular cohort groups or social contexts (i.e., subjective norms; Crant & Bateman, 1990). For example, in firms where management is responsible for imple- menting and monitoring the drug-testing program, it is likely that members of management will hold more favorable overall attitudes toward random drug testing than will nonmanagers. Therefore, it may be hypothesized that managers will perceive random drug testing to be less invasive than will nonmanagers (Hypothesis 2).

Individuals who are hired before drug testing is adopted should view the implementation of the program as a violation of operating norms, partic- ularly when the program calls for random drug testing. For these individ- uals, participation in drug testing may be viewed as an unanticipated cost of being employed by their firm. In contrast, individuals who are hired after drug testing is implemented and institutionalized should view random drug testing as accepted corporate policy and expect to be drug tested. Therefore, it may be hypothesized that individuals who are hired after random drug testing has been implemented will perceive the program to be less invasive than will individuals who are hired before the program is implemented (Hypothesis 3).

Program Characteristics Affecting Procedural Justice

This study focused on three characteristics of drug-testing programs that should have an impact on the experience of procedural justice: (a) the punitiveness of the consequences of drug detection, (b) the perceived

RANDOM DRUG TESTING 2 1 5

severity of the consequences of drug detection, and (c) the perceived accuracy of the drug-testing procedures. The following analysis outlines the manner in which each of these factors affects the experience of procedural justice and, hence, mitigates perceptions of the invasiveness of random- drug-testing programs.

Punitiveness of the consequences of drug detection. The conse- quences of drug detection range in punitiveness from reprimands to termination and, typically, firms employ more punitive consequences for multiple violations of the drug-testing policy (Axel, 1990). Consistent with procedural justice theory and research suggesting that individuals view less punitive procedures to be more ethical (cf. Leventhal, 1980), Stone and Kotch (1989) found that blue-collar workers held more favorable attitudes toward hypothetical drug-testing programs that involved rehabilitation and counseling rather than mandatory termination. Extending this research to the present context, it may be expected that individuals will find random drug testing to be less invasive when the policy gives those who test positive a "second chance" (i.e., employees are not terminated for testing positive for drug use on one occasion). Therefore, it may be hypothesized that employees who cannot be terminated for failing one drug test will perceive random drug testing to be less invasive than will individuals who can be terminated for failing one drug test (Hypothesis 4).

Perceived severity of the consequences of drug detection. Whether or not procedures are fair according to normative standards, perceptions of procedural justice are likely to engender favorable overall judgments (Lind & Tyler, 1988; Thibaut & Walker, 1975). In the context of drug-testing programs, it may be expected that individuals are likely to experience procedural justice when they perceive that the consequences of testing positive for drug use are not severe (whether or not the consequences are viewed as punitive in a normative sense; see Hypothesis 4). Hence, it may be expected that perceptions of the consequences of drug detection as less severe will mitigate the invasiveness engendered by random drug testing. Therefore, extending the arguments developed in Hypothesis 4, it may be hypothesized that individuals who perceive the consequences of drug detection to be less severe will find random drug testing to be less invasive than will individuals who perceive the consequences to be more severe (Hypothesis 5).

Perceived accuracy of drug-testing procedures. An additional de- terminant of procedural justice in the context of drug testing is the diagnostic capacity of the drug-testing procedures (i.e., the extent to which the drug-testing methods accurately discriminate users from nonusers).

2 1 6 TEPPER AND BRAUN

Consistent with Leventhal's (1980) accuracy rule of procedural justice, Gomez-Mejia and Balkin's (1987) study of human resource executives suggested that the diagnostic capacity or accuracy of the drug-testing procedures was related to the acceptability of drug testing. Notwithstanding the fact that many of the drug-testing methods currently used have serious measurement problems (from an objective standpoint; Bearman, 1988), accuracy perceptions may vary widely (cf. Gomez-Mejia & Balkin, 1987), and a conventional application of organizational justice theory requires examination of subjective procedural justice (cf. Lind & Tyler, 1988) from the perspective of those who undergo drug testing. Therefore, this field investigation tested the hypothesis that individuals who perceive the drug- testing procedures to be more accurate will perceive random drug testing to be less invasive than will individuals who perceive the procedures to be less accurate (Hypothesis 6).

METHOD

Design

The hypotheses were tested with employees from two firms headquartered in the east: a major air-carrier (employing over 20,(X)0) and a medium-sized utility (employing approximately 350). As mandated by federal guidelines, the two firms conducted random drug testing (with no advance notice given) on dl\ employees in safety-sensitive positions. Each of the drug-testing programs was approximately 8 years old, and they had changed little since they were implemented. The two firms tested their employees for drug use using radio immunoassay, and followed up all positive tests with a more sensitive method, the gas-chromatography/mass-spectrometry test.

Sampling and Survey Procedure

With the cooperation of the human resource departments at each firm, survey questionnaires were mailed to a random sample (selected from company records) of 100 employees at each firm. The target group consisted of technical/professional and managerial employees who were subject to random drug testing. The survey was printed on a business reply form and the respondents were instructed to mail their completed surveys back to the researchers. A cover letter guaranteed the respondents' ano- nymity and informed the participants that their employers would receive aggregate data summaries only. Of the 200 surveys mailed out, 122 were

RANDOM DRUG TESTING 2 1 7

returned (63 from Firm 1 and 59 from Firm 2; respondents with missing data were eliminated, producing a total sample size of 108 (58 from Firm 1 and 50 from Firm 2).

Respondent Characteristics

On average, the respondents from Firm 1 were 38.1 years old (SD = 8.56 years), 84.5Vo were male, 96.6<Vo were employed in various technical/ professional occupations (e.g., airline pilot, flight attendant, airplane mechanic, and customer service representative), and 2 participants de- scribed themselves as managers. On average, the respondents from Firm 2 were 41.1 years old (SD = 9.76 years), 9O<̂ o were male, and 2O<yo were employed in a managerial capacity. Eighty percent of the respondents from Firm 2 were employed in various technical/professional positions (e.g., chemical analyst, laboratory technician, and engineer), and 10 participants described themselves as manager.

Measures

Because the issues addressed by the survey were extremely sensitive, the firms required that the questionnaire be as short and innocuous as possible. Within those constraints, and the need to develop psychometrically sound measures for the hypothesis tests, a one-page survey questionnaire was constructed.

Criterion: Invasiveness perceptions. Three items, taken from Stone and Kotch (1989), assessed the respondents' judgments of the invasiveness of their drug-testing program: "Employee privacy is invaded by the way my company conducts drug testing," "The drug testing policy in my company is an invasion of privacy," and "I believe that my firm's drug testing policy violates employee rights." The 7-point response scale ranged from strongly disagree (1) to strongly agree (7). The alpha internal-consistency reliability coefficient for this measure was .93 in Firm 1 and .89 in Firm 2. The means for the participating firms were 12.19 (SD = 6.83) for Firm 1 and 9.38 (SD = 6.30) for Firm 2; these two means differed significantly, /(106) = 2.21,

p < .05, suggesting that the individuals from Firm 1 perceived their drug-testing program to be more invasive than did the individuals from Firm 2. However, removing the managers from the analysis revealed no differences between Firm i (n = 56, M = 12.32, SD = 6.84) and Firm 2 (n = 40, Af = 10.33, SD = 6.47), r(94) = 1.44, ns, suggesting that the differences in perceived invasiveness for the total sample may have arisen from the disproportionate number of managers who were tested for drug use in Firm 2.

2 1 8 TEPPER AND BRAUN

Predictors. Tenure was coded as follows: 1 = joined the firm after the drug testing policy had been implemented; 2 = member of the firm when drug testing was implemented. Managerial status was coded as follows: 1 = nonmanager; 2 = manager. The measure of the punitiveness of the consequences of drug detection asked the participants to indicate the consequences of testing positive for drug use once, twice, and three or more times. Scores were coded as follows: individuals who could be terminated for testing positive once were given a score of 1 (AI = 21 in Firm 1 and n = 1 in Firm 2), whereas those who could be terminated only after two or three positive drug tests received a score of 0 (n = 37 in Firm 1 and w = 43 in Firm 2).

Single-item measures assessed the employees' perceptions of the (a) diagnostic capacity of their firm's drug-testing procedures, (b) the severity of the consequences of drug detection, and the danger of impaired performance to (c) oneself, (d) one's co-workers, and (e) the public (scores on the last three items were summed to form an oversill measure of perceived danger). The perceptual items consisted of detailed definitions of the construct being assessed (i.e., accuracy, severity, or danger), followed by 9-point bipolar response scales. For accuracy, the scale ranged from very inaccurate (1) to very accurate (9); for severity, from not at all severe (1) to very severe (9) and for danger, from not at all dangerous (1) to very dangerous (9). The participants were instructed to circle the number (from 1 to 9) corresponding to their perceptions. The alpha internal-consistency reliability coefficient for the perceived danger measure was .83.

Control variables. Two variables likely to covary with invasiveness judgments, age and the number of times the respondent is tested for drug use, were measured in this study. According to Crant and Bateman (1989), older individuals are likely to have more favorable attitudes toward drug testing. Additionally, it may be expected that individuals who have under- gone more drug tests are likely to feel that the program represents a greater invasion of privacy than individuals who have undergone fewer drug tests.

Check For Response Bias

Because method variance can have a dramatic effect on relations between perceptual and affective measures, a check on this potential bias was undertaken. Harman's (1976) single-factor test was used to analyze the participants' responses to all the perceptual measures employed in this study (i.e., invasiveness judgments, perceived accuracy, perceived danger, and perceived severity) to determine whether a common-method factor might account for relations obtained between the predictors and the criterion. The existence of a common-method factor was not apparent in the unrotated

RANDOM DRUG TESTING 2 1 9

factor structure or in subsequent analyses of higher order factors. Hence, it does not appear that common-method bias produced artifactual relations between the perceptual variables employed in this study (Podsakoff & Organ, 1986).

RESULTS

Table 1 presents the means and intercorrelations for the study variables. Invasiveness perceptions correlated negatively with perceived accuracy (r = - .34, p < .01) and positively with perceived severity (r = .53, p < .01). Hence, employees perceived random drug testing to be less invasive when they perceived the drug-testing procedures to be more accurate and when they perceived the consequences of drug testing to be less punitive. Echoing the findings for perceived severity, the punitiveness of the consequences of drug detection correlated positively with invasiveness (r = .22, p < .05), suggesting that individuals who would receive a second chance after failing a drug test perceived random drug testing to be less invasive. Additionally,

TABLE 1 Descriptive Statistics and Intercorrelations for Study Variables

Correlations

Variable M SD 1 2 3 4 5 6 7 8

1. Invasiveness 10.89 6.71 - 2. Age 39.48 9.17 - . 1 3 3. Number of

times tested 1.84 1.40 .24* - . 2 7 * * 4. Tenure' 1.77 0.42 .16 .19* .14 5. Managerial

status" 1.11 0.32 - . 2 5 * * .27** - . 1 1 - . 0 2 - 6. Consequences" 0.26 0.44 .22* - . 1 6 .36** .07 - . 0 7 7. Perceived

severity** 5.77 3.03 .53** - . 1 3 .29** . 0 4 - 2 9 * * .30** 8. Perceived

danger' 21.91 6.92 .12 - . 1 3 .15 - . 0 7 - . 1 3 .00 .08 9. Perceived

accuracy^ 6.44 2.42 - . 3 4 * * .24* - . 0 2 - . 1 6 .10 - . 2 0 * - . 2 8 * * - . 0 5

' I = hired after drug-testing program was implemented; 2 = hired before drug testing was implemented. ''I = nonmanager; 2 = manager. "0 = individual could not be terminated after one positive drug test; 1 = individual could be terminated after one positive drug test. **Refers to the perceived severity of the consequences of drug testing. 'Refers to the perceived danger of impaired performance in one's own job. ^Refers to perceptions that the drug-testitig procedures accurately discriminate users from nonusers,

*p < .05. **p < .01.

2 2 0 TEPPER AND BRAUN

invasiveness correlated negatively with managerial status (r = - .25, /? < .01) and positively with the number of times tested for drug use (r = .24, p < .05), suggesting that managers viewed random drug testing to be less invasive than did nonmanagers, and individuals who had been tested for drug use more times viewed random drug testing to be more invasive than did individuals who had been tested fewer times. Age, tenure, and perceived danger were unrelated to invasiveness.

Although the bivariate relations are consistent with several of the hypotheses, more conservative analytical procedures were undertaken to determine the relative predictive power of the independent variables. Hierarchical regression equations were constructed in which the two control variables (age and the number of times tested for drug use) were entered first, the three distributive justice variables (tenure, managerial status, and perceived danger) were entered at the second step, and the three procedural justice variables (punitiveness of the consequences of drug detection, perceived severity of the consequences of drug detection, and the perceived accuracy of the drug-testing procedures) were entered at the third step. This order of entry follows the convention in the organizationail justice literature of examining the incremental contribution of procedural justice variables after controlling for the effects of distributive justice and control variables.

The hierarchical regression results shown in Table 2 indicate that the overall regression equation accounted for a substantial proportion of the variance in invasiveness perceptions: R^ = .35,/? < .01. Parenthetically, it should be noted that separate analyses for the two participating firms revealed findings similar to those obtained with the combined sample. The regression equations accounted for 33% (p < .01) of the variance for the first firm and 43% (p < .01) of the variance for the second firm. Moreover, applying the beta weights from the first firm to the variables in the second firm, and vice versa, produced only a modest drop in the variance accounted for by each equation: R^ = .22, p < .05, in the first firm and R^ = .32, p < .01, in the second firm.

As Table 2 shows, one of the control variables entered at the first step, the number of times tested for drug use, accounted for a significant proportionof the variance in invasiveness judgments: beta = .211 ̂ p < .05. The sign on the beta weight suggests that individuals who had been tested for drug use more times judged their program to be more invasive.

Only one of the distributive justice variables entered at Step 2 obtained a significant regression weight: managerial status, with a beta weight of - .212 (p < .05). The sign on the beta weight suggests that managers perceived their drug-testing program to be less invasive than did nonmanagers (supporting Hypothesis 2). Tenure and the perceived danger

RANDOM DRUG TESTING 2 2 1

TABLE 2 Hierarchical Regression Results

Step and Predictors

Step 1 Number of times tested** Age"

Step 2 Tenure'* Managerial status' Perceived danger*^

Step 3 Consequences of drug detection" Perceived accuracy'' Perceived severity'

Beta'

.217* - . 0 7 3

.148 - . 2 1 2 *

.065

.148 - .296**

.411**

Increment R^

06*

07*

22**

Total R^

06*

13*

35**

Note. For the full solution (three steps), FXS, 99) = 6.75, p < .01. 'Larger beta weights indicate greater invasiveness. ''Larger beta weights indicate greater

number of drug tests taken. '̂ Larger beta weights indicate older age. ^̂ l = hired after implementation; 2 = hired before implementation. '1 = nonmanager; 2 = manager. "̂ Larger beta weights indicate greater perceived danger. "Larger beta weights indicate more punitive consequences. ''Larger beta weights indicate greater perceived accuracy. 'Larger beta weights indicate greater perceived severity.

*p < .05. **p < .01.

of impaired job performance did not significantly affect invasiveness judgments. Hence, Hypothesis 1 and 3 were not supported.

At the third step, the punitiveness of the consequences of drug detection did not account for a significant proportion of the variance in invasiveness perceptions after entering the control variables and the distributive justice variables (at Steps 1 and 2, respectively). Hence, Hypothesis 4 was not supported. However, perceived severity and perceived accuracy obtained significant beta weights. The sign on the regression weight for perceived severity (beta = .411, /? < .01) suggests that employees who perceived the consequences of drug detection to be less severe perceived random drug testing to be less invasive (supporting Hypothesis 5). The negative sign on the regression weight for perceived accuracy (beta = - . 2 9 6 , p < .01) suggests that employees who perceived the drug-testing procedures to be more accurate perceived random drug testing to be less invasive (supporting Hypothesis 6).

DISCUSSION

This research extends the literature by examining actual employees* percep- tions of the invasion of privacy stemming from participation in random-

2 2 2 TEPPER AND BRAUN

drug-testing programs. Based on previous research suggesting that the experiences of distributive justice and procedural justice mitigate the negative effects of random drug testing on participants' attitudes, six determinants of justice were investigated. As predicted, drug testing was viewed as less invasive by managers, by individuals who perceived the consequences of testing positive to be less severe, and by individuals who perceived the drug testing procedures to be more accurate. Additionally, individuals who had been tested fewer times perceived drug testing to be less invasive.

The findings for managerial status are consistent with research suggesting that individuals are more committed to corporate policies that they are responsible for implementing and monitoring (cf. Staw, 1981). Future research should explore conditions under which managers' and employees' perceptions are likely to converge. For example, it may be expected that invasiveness perceptions are less divergent when drug testing is implemented at the request of the employees or when the terms of the policy are negotiated with union leadership (cf. LeRoy, 1991).

The results for perceived severity and perceived accuracy are encouraging because they suggest that the growing literature on procedural justice has implications for predicting individuals' attitudes toward drug testing in the workplace. The finding for perceived severity suggests that employee's subjective assessment of program punitiveness is related to their attitudes. The finding for perceived accuracy suggests that, even when the same methods of drug detection are employed across individuals, considerable variance in accuracy perceptions may be observed (see Table 1) and that these perceptions are likely to covary with perceptions of invasiveness. Additional research will be needed to determine how and why validity perceptions vary across individuals and what kinds of interventions may be useful in effectively informing employees of the accuracy of the procedures that most firms use.

Invasiveness perceptions were unrelated to age, tenure, the punitiveness of the consequences of drug detection, and the perceived danger of impaired job performance. The finding for perceived danger is perhaps the most surprising null result obtained here, although it confirms Murphy et al.'s (1991) speculation that general attitudes toward invasive drug-testing programs may be more favorable than individuals' attitudes toward their own drug-testing program.

From a practical standpoint, these findings suggest that implementation efforts should focus less on the safety-sensitive nature of jobs and more on individuals' perceptions of procedural justice. Particularly in firms that employ highly accurate (and costly) drug-testing procedures (e.g., gas chromatography/mass spectrometry) and in firms that have implemented comprehensive rehabilitation and employee assistance programs, morale

RANDOM DRUG TESTING 2 2 3

may be expected to increase if the work force is aware of the firm's commitment to enhancing procedural fairness (Greenberg, 1990a). Given the extent to which accuracy and severity perceptions may vary (whether or not the drug-testing procedures and the consequences of drug detection vary in an objective sense), firms can expect resentment to vary accordingly, particularly among nonmanagers.

Our study is not without limitations. By necessity, the perceptual measures of accuracy and severity were single-item scales. Additionally, the total sample consisted of only 108 individuals, even though the participating firms employed several thousand (in the case of the airline) and several hundred (in the case of the utility). It was encouraging, however, that cross-validation showed the findings to be reasonably robust across firms. Moreover, the fact that the two samples differed significantly in terms of invasiveness perceptions (see Table 1 and the earlier discussion) provides further evidence that the effects reported herein are neither context-specific nor artifacts of the research design.

It should also be noted that these methodological shortcomings reflect the difficulty of conducting psychometrically sound and theoretically mean- ingful field research in the drug-testing domain, particularly in firms that employ random-drug-testing procedures. Although many large companies may be optimistic about the success of their drug-testing program (Axel, 1990), researchers may encounter considerable opposition to the use of survey or interview methodologies when firms sense that such research may arouse employee resentment. It is hoped that these obstacles will not deter scholars from attempting to investigate drug-testing issues in the future. Additional, justice-based study of individuals from a broad sampling of occupations is needed to test other components of Crant and Bateman's (1989) model and to understand the processes that attenuate and augment the experience of organizational justice.

Questions about the generalizability of our findings should be addressed in future research. For example, additional research is needed to examine the generalizability of the present results to firms that freely choose to implement drug testing. It may be speculated that employees are more accepting of random drug testing when it is legally mandated (as in the firms studied here). Second, future research should assess the generalizability of the present findings to firms that test for cause, periodically, and as an initial condition of employment.

More practical concerns also suggest the need for further field research using a justice-based conceptual foundation. Legal scholars (e.g., Christopher, 1991) anticipate court challenges focusing on (a) the condi- tions under which employers may administer drug tests, (b) the drug-testing procedures that firms may employ, and (c) the range of acceptable consequences associated with drug detection. In addition, factors that have

2 2 4 TEPPER AND BRAUN

received little or no attention in the drug-testing literature (e.g., refusals to take drug tests, cheating on drug tests, and opportunities to appeal test results; Axel, 1990) warrant investigation. Our research suggests that theories of organizational justice provide a solid foundation for such research.

ACKNOWLEDGMENT

This research was supported by a grant from the Research Advisory Committee, University of Kentucky.

REFERENCES

Alexander, S., & Ruderman, M. (1987). The role of procedural and distributive justice in organizational behavior. Social Justice Research, 1, 177-198.

Axel, H. (1990). Corporate experiences with drug testing programs (Report No. 941). New York: The Conference Board.

Bearman, D. (1988). The medical case against drug testing. Harvard Business Review, 66, 126. Christopher, T. H. (1991). Employee drug testing: A constitutional perspective. Employee

Responsibilities and Rights Journal, 4, 311-328. Crant, J. M., & Bateman, T. S. (1989). A model of employee responses to drug-testing

programs. Employee Responsibilities and Rights Journal, 2, 173-190. Crant, J. M., & Bateman, T. S. (1990). An experimental test of the impact of drug testing

programs on potential job applicants' attitudes and intentions. Journal of Applied Psychology, 75, 127-131.

Folger, R., & Konovsky, M. A. (1989). Effects of procedural and distributive justice on reactions to pay raise decisions. Academy of Management Journal, 32, 115-130.

Gomez-Mejia, L. R., & Balkin, D. B. (1987). EHmensions and characteristics of personnel manager perceptions of effective drug-testing programs. Personnel Psychology, 40, 745-763.

Gordon, M. E., & Fryxell, G. E. (1989). Voluntariness of association as a moderator of the importance of procedural and distributive justice. Joumal of Applied Social Psychology, 19, 993-1009.

Greenberg, J. (1986). Determinants of perceived fairness of performance evaluations. Journal of Applied Psychology, 71, 340-342.

Greenberg, J. (1990a). Looking fair versus being fair: Managing impressions of organizational justice. Research in organizational behavior, 12, 111-157.

Greenberg, J. (1990b). Organizational justice: Yesterday, today, and tomorrow. Journal of Management, 16, 399-432.

Greenfield, P. A., Karren, R. J., & Giacobbe, J. K. (1989). Drug testing in the workplace: An overview of legal and philosophical issues. Employee Responsibilities and Rights Journal, 2, 1-10.

Harman, H. H. (1976). Modern factor analysis: 3rd edition. Chicago: University of Chicago Press.

Konovsky, M., & Cropanzano, R. (1991). Perceived fairness of employee drug testing as a predictor of employee attitudes and job performance. Journat of Applied Psychology, 76, 698-707.

RANDOM DRUG TESTING 2 2 5

LeRoy, M. H. (1991). Discriminating characteristics of union members' attitudes toward drug testing in the workplace. Journal of Labor Research, 12, 453-466.

Leventhal, G. S. (1980). What should be done with equity theory? New approaches to the study of fairness in social relationships. In K. Gergen, M. Greenberg, & R. Willis (Eds.), Social exchange: Advances in theory and research (pp. 27-55). New York: Plenum.

Lind, E. A., & Tyler, T. R. (1988). The social psychology of procedural justice. New York: Plenum.

Murphy, K. R., Thornton, G. C , & Prue, K. (1991). Influence of job characteristics on the acceptability of employee drug testing. Journal of Applied Psychology, 76, 447-453.

Murphy, K. R., Thornton, G. C , «& Reynolds, D. H. (1990). College students' attitudes toward employee drug testing. Personnel Psychology, 43, 615-631.

Podsakoff, P. M., & Organ, D. W. (1986). Self-reports in organizational research: Problems and prosperts. Journal of Management, 12, 531-544.

Staw, B. M. (1981). The escalation of commitment to a course of action. Academy of Management Review, 6, 577-587.

Stone, D. L., & Kotch, D. A. (1989). Individuals' attitudes toweu-d organizational drug testing policies and practices. Journal of Applied Psychology, 74, 518-521.

Thibaut, J., & Walker, L. (1975). Procedural justice: A psychological analysis. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.