Journal of Applied Psychology Copyright 1992 by the American Psychological Association, Inc.

1992. Vol. 77. No. 3. 288-297 ' 0021-9010/92/S3.00

The Effects of Met Expectations on Newcomer Attitudes

and Behaviors: A Review and Meta-Analysis
John P. Wanous and Timothy D. Poland
College of Business
The Ohio State University
Stephen L. Premack and K. Shannon Davis
College of Business
University of Maryland, College Park

A review of research on the effects of met expectations for newcomers to organizations located 31
studies of 17,241 people. A meta-analysis found mean (corrected) correlations of.39 for job satisfac-
tion and for organizational commitment, .29 for intent to leave,. 19 for job survival, and. 11 for job
performance. However, all of these mean correlations had significant between-studies variance. By
using strict conformity with Porter and Steers's (1973) definition of met expectations, we identified
a subset of studies that had nonsignificant between-studies variance for all correlations except job
satisfaction. Furthermore, the mean correlations in these subgroups were very similar to those for
the entire group. Future research should consider both the direction of the met expectations dis-
crepancy (i.e., over- vs. underfulfillment) and alternative ways to measure organizational reality.

The concept of met expectations in the research literature of because (a) the expectations held by new recruits are almost
industrial and organizational psychology and organizational always inflated (Wanous, 1980,1992), (b) turnover rates among
behavior (I/OB) has been mentioned frequently for over 30 newly hired employees are typically much higher than among
years, but less frequently studied. In the first published experi- employees with greater tenure in an organization (Mobley,
ment on realistic job previews, Weitz (1956) alerted practi- 1982), and (c) RJPs have been shown to lower expectations and
tioners and researchers alike to the potential usefulness of hav- modestly increase job survival (Premack & Wanous, 1985).
ing employee expectations be as realistic as possible. In a widely Given this, it is not surprising that the importance of met ex-
cited review of research on employee turnover, Porter and pectations has been accepted despite the lack of a systematic
Steers (1973) articulated the met expectations hypothesis as it is research review.
known today in I/OB: Met expectations has also been a key psychological variable
The concept of met expectations may be viewed as the discrep- in research on the effectiveness of different recruiting sources
ancy between what a person encounters on the job in the way of (Wanous & Colella, 1989). Briefly, met expectations is one expla-
positive and negative experiences and what he expected to en- nation of why certain inside sources, such as rehired employees
counter. Thus, since different employees can have quite different or employee referrals, result in higher job survival rates than do
expectations with respect to payoffs or rewards in a given organiza-
tional or work situation, it would not be anticipated that a given outside sources such as newspaper ads or employment agencies.
variable (e.g., high pay, unfriendly work colleagues, etc.) would This is because it is assumed that inside sources provide more
have a uniform impact on withdrawal decisions. We would pre- accurate information about a particular organization, acting
dict, however, that when an individual's expectations—whatever somewhat like an RJP. A recent review of 12 studies of recruit-
they are—are not substantially met, his propensity to withdraw
would increase. (Porter & Steers, 1973, p. 152) ing source effectiveness (Wanous, 1992) found that inside
sources had job survival rates that were about 30% higher than
Reviews of the realistic job preview (RJP) literature (Premack those of outside sources. (When the 12 studies were weighted by-
& Wanous, 1985; Wanous, 1977,1980,1992; Wanous & Colella, sample size, inside sources had a 24% higher survival rate; when
1989) have kept the topic alive because RJPs represent one way the studies were weighted equally, the inside source survival
of creating met expectations. rate was 36% higher. Equal weighting was done as a comparison
Despite several reviews of both the turnover and RJP litera- because 2 of the 12 studies accounted for 70% of the total sam-
tures, no previous review has examined the research that has ple size.)
directly focused on the met expectations hypothesis. Some text- Met expectations has also been an important psychological
books on selection and staffing (e.g., Schneider & Schmitt, variable in various stage model theories of organizational so-
1986) have accepted the validity of this hypothesis, probably cialization (e.g., Feldman, 1976; Porter, Lawler, & Hackman,
1975; Schein, 1978: Van Maanen, 1976; Wanous, 1980). Basi-
cally, all of these models assume that unmet expectations cause
Stephen L. Premack is now deceased.
Correspondence concerning this article should be addressed to John a variety of postentry adjustment problems, for example, low
P. Wanous, College of Business, The Ohio State University, Columbus, job satisfaction and early turnover.
Ohio 43210. Research on met expectations in the I/OB area is a specific
288
 EFFECTS OF MET EXPECTATIONS 289

example of two concepts found in several other bodies of re- cerns the specific meaning of met expectations. In this case, a
search: (a) the role of expectations in motivation, decision mak- discrepancy is assessed between one's initial expectations and
ing, or general cognitive activity, and (b) the concept of match- one's subsequent beliefs after entering an organization and expe-
ing, congruence, or fit. For example, the importance of expec- riencing it as a full-time member. The operational definition of
tations in management decision making was explicitly this discrepancy and the appropriate statistical analysis are,
acknowledged by Cyert and March (1963) and by many others. however, areas of controversy, which we discuss later.
Expectations in theories of work motivation were considered The fourth aspect of Porter and Steers's (1973) definition con-
crucial in Vroom's (1964) version of expectancy theory and in cerns the meaning of expectations. Only those expectations for
all subsequent formulations of it. For example, when applying important aspects of the job or organization are included in the
expectancy theory to choosing among job offers, Vroom used met expectations hypothesis, not all expectations per se, be-
instrumentality to refer to the expectation of certain outcomes cause that would include irrelevant or inconsequential expecta-
that would occur if one joined a particular organization. The tions. Porter and Steers were not as explicit about this distinc-
typical measure of this type of expectation is a set of items in tion as Locke (1976) was a few years later. However, a careful
which respondents are asked to rate the likelihood of certain reading of Porter and Steers's work reveals that they considered
outcomes being present in a particular organization (Wanous, only the disconfirmation of important expectations to be dis-
Keon, & Latack, 1983). Outside the I/OB area, met expecta- satisfying.
tions has been a topic of considerable research by social psychol- As can be seen from the preceding discussion, Porter and
ogists concerned with cognitive dissonance (see Abelson et al., Steers's (1973) definition of met expectations has a rather spe-
1968) and, more recently, by experimental psychologists con- cific meaning. The four implied facets of the definition served
cerned with stress in aversive situations (e.g., Abbott & Badia, as an initial set of criteria for our selection of the research to be
1986; Abbott, Schoen, & Badia, 1984; Arthur, 1986). These few reviewed and included in the current meta-analysis. However,
examples are not intended to be exhaustive, but rather illustra- the number of studies that used this precise definition was
tive of how pervasive expectations are throughout I/OB and small, so we relaxed the criteria somewhat to locate a larger, but
related areas of research. still relevant, body of research.
Similarly, one can find wide-ranging examples of research Specifically, studies that related met expectations to one or
concerned with the concept of matching, congruence, or fit. more of several attitudes (job satisfaction, organizational com-
For example, most of the concern with staffing organizations mitment, and the intention to remain) or behaviors (job sur-
involves getting appropriate matches between job candidates' vival and job performance) were included, even if Porter and
capabilities and organizational requirements on the one hand Steers's exact causal sequence had not been studied. Similarly,
and job candidates' wants or needs and organizational climates some correlational and laboratory studies were included if they
or cultures on the other (Caldwell & O'Reilly, 1990; Schneider & concerned expectations about jobs and organizations. Studies
Schmitt, 1986; Wanous, 1980, 1992). Similarly, leadership re- that used non-discrepancy-score measures of met expectations
search has considered the fit of leadership style to the type of were also included.
decision (Vroom & Yetton, 1973) and the fit of the type of
leader to the situation (Fiedler, 1967; Fiedler & Garcia, 1987).
Research on stress has examined person-environment fit as Method
well (Edwards & Cooper, 1990). Selection of Studies and Coding of Study Characteristics
Definition of Met Expectations in Industrial and Four criteria were used for study selection. First, an individual's
expectations about work-related conditions (e.g., pay, supervision, etc.)
Organizational Psychology and Organizational must have been measured. Second, the sample size must have been
Behavior Research reported. Third, a Pearson correlation coefficient (or some other type
The definition of met expectations in I/OB research comes of statistic, e.g., a t or F ratio that could be converted into a correlation)
from the work of Porter and Steers (1973). The first aspect of between met expectations and at least one of the five outcome vari-
ables must have been reported. Fourth, the study had to measure ex-
Porter and Steers's definition is the basic hypothesis itself. Un- pectations directly. This meant that only 6 of the 20 RJP experiments
met expectations are seen as leading to dissatisfaction, which in reviewed by Premack and Wanous (1985) could be included here, be-
turn leads to quitting an organization. Thus, two links are speci- cause the other 14 did not report correlations between met expecta-
fied in the hypothesis, in which satisfaction mediates the rela- tions and the five outcomes investigated here.
tionship between unmet expectations and quitting (or job sur- Studies were located by manual searches of PsycSCAN:Applied Psy-
vival). chology, Dissertation Abstracts, and Dissertation Abstracts Interna-
The second aspect of Porter and Steers's (1973) definition tional, and by scanning the reference lists of published and unpub-
concerns the appropriate context for conducting research. In lished sources. The 20 studies cited by Cotton and Tuttle (1986) were
this case, expectations held by job candidates before they enter examined, but only 6 met the criteria used here. Our literature search
an organization are compared with their postentry experiences. produced 31 studies that met the four criteria. Seven of these studies
are unpublished, which minimizes the "file drawer problem" (Ro-
This also means that the relevant expectations concern both the senthal, 1979). A total of 17,241 individuals participated in the 31
specific job and the wider organizational context. Thus, met studies; the mean sample size was 556.
expectations research should be conducted with job candidates After identification of the studies to be used in the analysis, each
who later become new recruits. study was coded on several factors (see the Appendix). All studies were
The third aspect of Porter and Steers's (1973) definition con- coded twice: once by Stephen L. Premack and again by K. Shannon
290 WANOUS, POLAND, PREMACK, AND DAVIS

Davis; John P. Wanous and Timothy D. Poland double-checked the pectations, job satisfaction, and organizational commitment, but
coding. Before any discussion between coders, intercoder agreement never for job performance or job survival. In several studies, correla-
for the various information taken from the studies exceeded 90% for tions were reported for more than one time period, and in one case
all variables. The few cases involving disagreement were resolved by they were broken out by the type of organization. When this occurred,
subsequent discussion, and complete agreement was reached in all correlations were averaged. No attempt was made to correct for restric-
cases. Interestingly, the 10% of cases on which coders disagreed were tion in range because the data that would have made this correction
due entirely to a misreading of a study by one coder, rather than to two possible (population means and standard deviations) were unavailable.
different judgments. Thus, we believe that the studies are accurately The last correction was done on the job survival data. This is be-
coded. Nevertheless, using multiple coders was considered important cause the relationship between met expectations and job survival is
because research on meta-analysis procedures has shown that human reported as a point biserial correlation. We consider this to be a form of
judgment calls can affect the results (Wanous, Sullivan, & Malinak, artifactual dichotomization, given that the underlying construct of
1989). tenure or participation is continuous (March & Simon, 1 958; Porter &
Because the definition of met expectations was an important source Steers, 1973). When this occurs, the corrected correlation is the biser-
of variation among the studies, the operational measure of expecta- ial correlation (Hunter & Schmidt, 1990, p. 335). The formula used to
tions from each study was independently coded by five other judges. correct the correlations is
The judges were organizational behavior or personnel and human re-
source management faculty members. Each judge was provided with a
coding sheet that presented Porter and Steers's (1973) definition of met
expectations and was asked to indicate whether the expectations mea-
sure was consistent with Porter and Steers's definition or represented where /t is the biserial correlation, r^ is the point biserial correlation, p
some other definition (a dichotomous judgment). On the basis of this is the proportion of stayers, q is the proportion of leavers, and h is the
criterion, 18 studies were found that used operational measures consis- ordinate of the unit normal distribution at p (Williams, 1990, p. 733).
tent with Porter and Steers's conceptualization; the 13 other studies There is some controversy surrounding whether or not such a correc-
used different definitions. tion should be made and, if so, which procedure should be used (see
Five correlates of met expectations were investigated frequently Bass & Ager, 1991; Williams, 1990). We report the results both with
enough to be included in the meta-analysis; that is, at least 3 studies and without this correction.
reported data for a particular outcome (see Pearlman, Schmidt, &
Hunter, 1980, for further information on inclusion criteria). The actual
number of studies ranged from 10 to 19 depending on the particular
Results
correlate of met expectations. Each correlate of met expectations is
described in the following paragraphs:
1. Organizational commitment was measured with the Organiza- Table 1 shows the results for five correlates of met expecta-
tional Commitment Questionnaire (OCQ; Mowday, Steers, & Porter, tions. The average correlation and the corrected average corre-
1979) in all studies. lation (for attenuation due to unreliability) are shown first,
2. Intention to remain was typically measured with a single item that along with a 95% confidence interval around each. These are
asked employees their intention to quit (reverse scored) or remain. followed by the results of the meta-analysis as calculated from
3. Job satisfaction was measured in several ways, ranging from ad the dstatistic (Hunter & Schmidt, 1990). The chi-square test for
hoc items to better known scales, such as the Minnesota Satisfaction the significance of between-studies variance is shown last
Questionnaire or the Job Descriptive Index.
(Hunter & Schmidt, 1990, p. 428). The null hypothesis for this
4. Job survival was typically measured (in 16 of 18 studies) as a di-
chotomous variable (i.e., as stay vs. leave). In some cases, the sign of a test is that there is no between-studies variance in the mean
correlation between met expectations and job survival was reversed to effect size (Hunter & Schmidt, 1990, p. 213). If this null hypoth-
be consistent with the hypothesis that met expectations increase job esis is rejected, then one cannot conclude for certain that a true
survival. population mean has been found. That is, there is too much
5. Job performance was measured in a variety of ways, ranging from unexplained between-studies variance in the effect size, even
supervisory and self-ratings of performance to the quality and quantity after removing the variance due to sampling error. On the other
of output. hand, if the null hypothesis is not rejected, then one can con-
clude that a true population mean has been found. That is, after
between-studies variance attributable to sampling error is ac-
Meta-Analysis Procedure counted for by meta-analysis, the remaining variance among
Because reliability data were not reported for every study investigat- effect sizes is not significantly different from zero. When this
ing a particular outcome, correction for attenuation due to unreliabil- occurs, confidence intervals are omitted.
ity was performed across studies. Thus, correlations were first sub- The table also shows the results of a moderator analysis in
jected to a meta-analysis to eliminate the effects of sampling error, which the studies were subgrouped according to their confor-
followed by a correction of the mean correlation by the mean of the mity with Porter and Steers's (1973) definition of met expecta-
reported reliabilities (Hunter & Schmidt, 1990). In three studies tions. This was done because the initial meta-analysis done on
(Dean, 1981; Greenhaus, Seidel, & Marinis, 1983; Homer, 1979), reli- all five correlates found significant amounts of between-studies
ability data for the met expectations measure or the job satisfaction variance, even after correcting for sampling error.
measure were reported for subscales rather than the complete scale. In
The definition of met expectations (Porter and Steers's vs.
this case, the mean subscale reliability coefficient was used, but only
after it was corrected by the Spearman-Brown formula. This was done others') was our first choice for a logical moderator variable
because the likely lower reliabilities of subscales could be a source of because conformity to the operational definition implied by
between-studies variance. Porter and Steers seemed very basic. We reasoned that studies
Reliability data were sometimes available for measures of met ex- not conforming to Porter and Steers's definition would be likely
 EFFECTS OF MET EXPECTATIONS 291

Table 1
Meta-Analyses of the Effect of Met Expectations on Newcomer Attitudes and Behavior
95% 95%
No. of confidence confidence
Attitude or behavior/studies* studies N ,. interval 'cor interval d y
y. y
df x2
Job satisfaction
All 19 3,960 .33 .051 -.602 .39 .061-.725 .72 .132 .020 .112 18 122.96**
Porter and Steers's (1973) definition
4, 8, 1 1, 12, 14, 15, 16, 19, 22, 23, 28, 29 12 2,444 .33 .023-.601 .36 .027-.702 .68 .132 .021 .112 11 76.39**
Other definition
1,7, 10, 18,20,24,27 7 1,516 .35 .104-. 596 .45 .133-.764 .77 .127 .020 .107 6 44.78**
Porter and Steers's definition
(outliers removed)
4,8, 11, 12, 14, 16, 19,23,28,29 10 2,142 .28 .033-.524 .32 .038-.606 .60 .088 .020 .068 9 44.90**
Organizational commitment
All 15 2,991 .33 .157-.494 .39 .187-.590 .70 .052 .021 .031 14 36.51**
Porter and Steers's definition
3,4,5,6,8, 11, 14, 19,21,29 10 1,796 .29 .34 — .61 .015 .023 -.009 9 6.25
Other definition —
1,20,24,25,27 5 1,195 .38 .188-.568 .45 .22S-.679 .83 .077 .018 .059 4 21.12**
Intent to remain
All 14 2,851 .24 .072-.399 .29 .089-.493 .49 .038 .020 .018 13 26.41*
Porter and Steers's definition
4,5,6,8, 11, 14, 19,21,29 10 1,924 .24 .28 .50 .016 .021 -.005 9 7.51
Other definition — —
1,20,24,27 4 927 .23 -.066-.457 .27 -.066-. 540 .48 .084 .018 .066 3 18.87**
Job performance
All 10 2,130 .09 -.105-.282 .11 -.126-.340 .18 .046 .019 .027 9 24.26**
Porter and Steers's definition
3,4,5, 11, 14, 16, 19 7 1,259 .10 .12 .20 .017 .022 -.005 6 5.41
Other definition — —
18,26,27 3 871 .07 -.184-.331 .09 -.223-.40I .16 .086 .014 .072 2 18.62**
Job survival
All 18 14,210 .12 -.034-.269 .13 -.038-.300 .24 .037 .005 .032 17 129.34**
Porter and Steers's definition
5,8, 11, 14, 17,21,28,30 8 3,267 .16 -.094-.416 .17 -.101-.446 .34 .101 .010 .091 7 81.57**
Other definition
1,2,7,9, 13,20,25,26,27,31 10 10,943 .10 .018-.192 .12 .021 -.225 .21 .013 .003 .010 9 36.63**
Porter and Steers's definition
(outlier removed)
5,8, 11, 14, 17,21,28 7 3,003 .13 .14 .25 .014 .009 .004 6 10.34
Job survival (corrected for dichotomization)
All 16 13,554 .17 -.017-.355 .19 -.019-.396 .36 .073 .005 .068 15 242.56**
Porter and Steers's definition
5,8, 11, 14, 17,21,28,30 8 3,267 .23 -.113-.570 .24 -.121-.611 .52 .256 .010 .246 7 201.96**
Other definition
1,2,7, 13,20,25,27,31 8 10,287 .15 .18 .30 .0033 .0031 .0002 7 8.39**
Porter and Steers's definition — —
(outlier removed)
5,8, 11, 14, 17,21,28 7 3,003 .19 -.013-.384 .20 -.014-.412 .39 .054 .009 .045 6 40.11**
Note. Confidence intervals were not calculated when between-studies variance was nonsignificant. Obs = observed; cor = corrected; e = error;
pop = population.
* See the Appendix for the full citations of the numbered studies.
*p<.05. **p<.0l.

to have more between-studies variance among themselves as a could have the effect of increasing the number of met expecta-
subgroup than the subgroup of studies that did conform to tions for half of the people in the study (the experimental
Porter and Steers's definition. group). Interestingly, if the RJP experiment succeeded in doing
We tried two other potential moderators: (a) whether the precisely this, it would also have the effect of reducing the vari-
study was experimental or nonexperimental in design, and (b) ance in met expectations for those in the experimental group,
whether the study was published or not. Seven of the eight exper- which could attenuate the correlation between met expecta-
imental studies were from RJP experiments in which the ma- tions and various outcomes. This is because the experimental
nipulation of expectations was central to the research. If the group is combined with the control group for the purpose of
RJP treatment successfully manipulated expectations, this correlating met expectations with various outcomes. Thus, it
292 WANOUS, POLAND, PREMACK, AND DAVIS

seemed possible that experimental studies might have lower what one would expect. That is, the strongest results are found
correlations than nonexperimental studies. for job satisfaction and organizational commitment, followed
The other potential moderator was tried because much has by intent to remain and, last, job survival. This is the type of
been written on the possible differences between published causal sequence suggested in the met expectations literature.
and unpublished research. Despite the reasonableness of se- The fact that the mean correlations decrease is logical, because
lecting these two moderators, neither accounted for any addi- the outcome variables are increasingly distant from their pur-
tional significant between-studies variance in the effect sizes. ported cause (met expectations). Thus, job survival should have
The results for job satisfaction are shown at the top of Table 1. a weaker relationship to met expectations than to initial atti-
Despite the relatively strong corrected correlation of .39, the tudes (satisfaction and commitment). The fact that intent to
between-studies variance was significant no matter which type remain was in between attitudes and behavior (in terms of the
of moderator analysis was tried. Even when two outlier studies strength of the relationship) is also consistent with decades of
were removed from the Porter and Steers group, the corrected research on the relationship between attitudes, intentions, and
variance was still significant for the Porter and Steers group. behavior (McGuire, 1985).
The mean correlations for organizational commitment were The fact that met expectations correlates weakest with job
identical with those for job satisfaction. However, in the moder- performance might be considered as a form of discriminant
ator analysis, the Porter and Steers group had a nonsignificant validity. This is because job performance was never suggested
amount of corrected between-studies variance, but it also had a as a likely outcome of met expectations by Porter and Steers.
lower mean correlation than did the Other group (.34 vs. .45). There was, nevertheless, a modest relationship between met
For intent to remain, the Porter and Steers versus Other cate- expectations and performance, as suggested by the results for
gorization was once again a useful moderator because the the Porter and Steers subgroup (r = .12). This link to perfor-
corrected variance for the Porter and Steers group was nonsig- mance may also make sense. For example, it has been suggested
nificant. There was little difference in the mean correlations by those concerned with RJPs (Premack & Wanous, 1985;
between these two subgroups (.28 vs. .27). Wanous, 1992) that performance can be increased by met ex-
The average correlation for job performance was the weakest pectations if the job preview includes information that clarifies
of any outcome measure. The moderator analysis again showed role expectations. Furthermore, job performance has been
that the Porter and Steers group had a nonsignificant between- shown to have a modest relationship with job satisfaction (laf-
studies corrected variance (and a slightly stronger average faldano & Muchinsky, 1985; Petty, McGee, & Cavender, 1984).
correlation) than did the Other group (i.e., .12 vs. .09). Because of this, some of the increased performance from met
The moderator analysis for job survival required that one expectations might also be due to the effect of met expectations
outlier study be dropped from the Porter and Steers group for on job satisfaction. Thus, it is reasonable that there was some
the between-studies corrected variance to be nonsignificant. relationship with job performance, but it is also reasonable that
The outlier was an unpublished doctoral dissertation (Young- the relationship was weaker than for the more theoretically
berg, 1963) that reported data much stronger than the others in relevant outcomes.
this group, and it greatly contributed to the between-studies If the pattern of results seems to make sense theoretically,
variance. This study was also considered an outlier in a meta- does the level of results also make sense? That is, are the mean
analysis of RJP experiments (see Premack & Wanous, 1985). correlations interpretable within existing frameworks? One way
When the job survival correlations are corrected fordichoto- to address this is to compare these results with those from
mization, the number of studies dropped from 18 to 16 because RJPs, as these are somewhat related bodies of research and do
of incomplete information necessary to conduct the correction have some studies in common. Because met expectations have
for dichotomization. The effect size increased by 50% as a result been suggested as one of the reasons why RJPs increase job
of the correction. The Porter and Steers subgroup of studies had survival, it is logical to expect that they would have a stronger
significant between-studies variance, which it did not before relationship with job survival than would the RJP per se. This is
this correction was made. The Other subgroup had nonsignifi- because met expectations are the result of an RJP and, thus, are
cant between-studies variance, just the reverse of the situation closer to the behavior they are supposed to cause (job survival).
before the correction was made. The average correlation between met expectations and job sur-
One final word on the moderator analysis concerns the re- vival was found to be. 14 (for the Porter and Steers group), and
classifying of outlier studies. A number of statistics experts this can be compared with the average correlation of .06 be-
have recommended that data should always be examined for tween the RJP and job survival reported by Premack and Wan-
the effects of outliers (Cook, 1977; Stevens, 1984; Tukey, 1977). ous (1985).
This is especially incumbent upon the researcher when the sam- Calculating the increase in job survival rates can be done in
ple size is relatively small. A recent article on this topic con- the same way as suggested by Premack and Wanous (1985, p.
cluded that researchers should examine outlier effects and re- 716). To begin with, the mean effect size of .25 for the outlier-
port them explicitly (Orr, Sackett, & Dubois, 1991). We have removed Porter and Steers group is probably the best result to
done this here. use for this calculation, because there is no significant between-
studies variance. This effect size means that those employees
whose expectations are met will have a job survival rate that is
Discussion .25 standard deviation units greater than the job survival rate
From the theoretical perspective of met expectations as origi- for those whose expectations are not met. If the job survival rate
nally formulated, the pattern of average correlations is about is .50 (for, say, the first year on the job), then the expected job
 EFFECTS OF MET EXPECTATIONS 293

survival rate would be .625 if met expectations could be created effect size, adding to its likely generalizability beyond just the
for all new employees. (This is because the standard deviation Porter and Steers subgroup of studies.
of a .50 survival rate is .50. Thus, the gain in job survival is .25 X Even if the results reported in Table 1 seem to make sense
.50 =.125). theoretically, it is still possible that they have been produced
The effect of met expectations on job satisfaction should be artifactually by the methods used here. Critics of meta-analysis
stronger than the effect of an RJP on job satisfaction. This is (e.g., Spector & Levine, 1987) have noted that having a small
because job satisfaction is assumed to be immediately affected number of studies can lead one to conclude that there is no
in met expectations theory, whereas it is hypothesized to be a between-studies variance when, in fact, there really is. In other
more distant effect of an RJP (Wanous, 1980,1992). The results words, critics contend that Hunter and Schmidt's (1990)
of research confirm that RJPs have much less effect on job method biases one to believe that results are generalizable
satisfaction (average correlation of .05) than do met expecta- when they really are not. This criticism would seem to apply
tions (average correlation of .39). However, one reason why met here because the number of studies is small, at least when com-
expectations might appear to have the stronger relationship is pared with selection-test meta-analyses (Hunter & Hunter,
because there is common method variance between met expec- 1984; Schmitt, Gooding, Noe, & Kirsch, 1984). However, re-
tations and job satisfaction (i.e., both are measured on question- cent articles have detected a flaw in Spector and Levine's (1987)
naires). study (see Callender & Osburn, 1988; Rasmussen & Loher,
1988). Rasmussen and Loher's (1988) study is particularly rele-
Even though comparisons have been made between the re-
vant because it shows that Type I error rates are not at all exces-
sults for met expectations and those for RJPs, one must be
sive, even when there are as few as 6 or 10 studies and the
cautious in doing so. For the most part, the two bodies of re-
underlying population correlation is .30. These parameters are
search are not the same. Specifically, only 7 of the 31 studies
almost identical to the research on met expectations reviewed
reported here are RJP experiments, and only 6 of these were
here.
included in Premack and Wanous's (1985) meta-analysis, which
Although this most recent work shows that Type I errors are
has been used here for comparison purposes.
not excessively high when the number of studies is small, it does
A final note on the theoretical meaningfulness of this meta- indicate that the power to detect the presence of a moderator is
analysis is in order here. A review of Table 1 shows that the unacceptably low (Kemery, Mossholder, & Dunlap, 1989; Ras-
meta-analysis was successful in explaining the variance in the mussen & Loher, 1988). However, this criticism is irrelevant
effect of met expectations on organizational commitment, in- here because we were able to detect a moderator in most cases.
tent to remain, job performance, and job survival. However, it The results reported here also tend to reduce concerns about
was unsuccessful in explaining the variance in the effect of met the number of studies reviewed. This is because the mean
expectations on job satisfaction. A likely explanation for this correlations in the Porter and Steers and Other subgroups are
result lies with the operationalization of the criterion variables. very similar to each other. Take job survival as an example.
As mentioned earlier, organizational commitment was always What seems to have happened is that there are more extreme
measured with the OCQ (Mowday et al., 1979), whereas job values in the Other group, because the average corrected corre-
satisfaction was measured in a variety of different ways. The lation (.14) is the same for both groups in the uncorrected analy-
fact that we were able to account for the between-studies vari- sis and almost identical (.20 vs.. 18) in the corrected analysis. A
ance in the effect size of met expectations on organizational similar pattern was found for intent to remain and for job per-
commitment should not be surprising because both the predic- formance. For these three correlates of met expectations, then,
tor measure (Porter and Steers's, 1973, definition of met expec- it may be reasonable to believe that the population correlation
tations) and the criterion measure (the OCQ) were held con- has been discovered. This is because the overall mean correla-
stant across studies. On the other hand, the unexplained be- tion is the same as that for the Porter and Steers subgroup.
tween-studies variance in the effect of met expectations on job Because the overall correlation is based on more studies than
satisfaction might be due to the variability in the measurement are found in the Porter and Steers subgroup, one can have more
of job satisfaction across studies. confidence in the results. There were small differences between
The results for job survival deserve special mention. The use the Porter and Steers and Other subgroups for job satisfaction
of the Porter and Steers versus Other distinction as a moderator and organizational commitment, so that it is more conservative
produced different results, depending on whether a correction to interpret the population correlation as being represented by
was made for dichotomization. Without the correction, the the Porter and Steers subgroup, rather than the overall result.
Porter and Steers group had nonsignificant between-studies
variance, making the mean effect size directly interpretable. Future Research
However, when the correction was made, the Other group had
the nonsignificant between-studies variance. This makes it dif- The suggested future research topics come from a general
ficult to say which mean correlation is the one that should be reading of the met expectations literature. Thus, they are not
discussed. The most conservative approach is to focus on the necessarily derived from the means and variances of the meta-
outlier-removed Porter and Steers group (uncorrected for di- analyses.
chotomization) because the between-studies variance was non-
Direction of Discrepancies
significant and because this is the group of studies meeting the
original definition of met expectations. Furthermore, this ef- One topic omitted by researchers interested in met expecta-
fect size (d= .25) is virtually identical to the overall uncorrected tions is the degree to which getting less than expected is the
294 WANOUS, POLAND, PREMACK, AND DAVIS

equivalent of getting more than expected. Porter and Steers port the separate and combined effects of expectations and per-
(1973) suggested that low levels of expectations are desirable ceptions.
because they will have a higher probability of being fulfilled. In several other studies, between-person discrepancy scores,
However, Porter and Steers did not discuss what might happen rather than within-person discrepancy scores, were used. By
if a person's expectations are unmet in the sense of being over- between-person, we mean that the newcomer's expectations
fulfilled. were compared with someone else's perception of organiza-
Equity theory (Adams, 1963) is a good example of explicit tional reality. In a strictly psychometric sense, between-person
discussion of overfulfillment. In fact, over- versus underpay- discrepancy measures are less error prone than within-person
ment has been a topic of considerable interest (Campbell & measures (Johns, 1981). In a theoretical sense, however, a be-
Pritchard, 1976; Lawler, 1971). Equity theory assumes that get- tween-person discrepancy measure implies a different concept
ting more than is considered fair leads to some type of psycho- of organizational reality. This is because the comparison stan-
logical tension, resulting in dissatisfaction and actions to rem- dard is a constant value for all newcomers (e.g., the mean of a
edy the dissatisfaction. Empirical research on equity theory, person's work group perceptions, or a boss's perception of
however, shows that over- versus underreward do not have sym- reality).
metrical effects—the threshold for overreward is higher than We found at least four different between-person discrepancy
for underreward (Campbell & Pritchard, 1976; Lawler, 1971). measures of met expectations. First, newcomer expectations
Although met expectations research has not explicitly ac- have been compared with the mean of insider perceptions, as
counted for the direction of the discrepancy, research on the was done by Schneider (1975), who compared newcomers' ex-
measurement and meaning of job satisfaction (Wanous & pectations with the mean of their work groups' perceptions—
Lawler, 1972) has examined the issue recently (Rice, McFarlin, in this case, employees who worked together in an insurance
& Bennett, 1989). However, the job satisfaction research is con- agency. Second, a variation on comparisons with a group mean
cerned with a different standard of comparison to one's job was proposed by Mowday (1980), who considered an expecta-
experiences (i.e., what is wanted vs. what is expected). This dif- tion to be unmet only if it deviated from the group mean by
ference could be important because receiving more than ex- more than one standard deviation (in either direction). Third,
pected may not necessarily cause dissatisfaction if the specific newcomer expectations have been compared with the percep-
factor is one that is highly desirable or valent (Vroom, 1964). tion of reality by an influential person in the newcomer's role
set (e.g., the newcomer's own boss; Schneider, 1975). Fourth, a
group of knowledgeable insiders (not necessarily those in face-
What Is Reality to-face contact with a newcomer) were asked to describe organi-
zational reality individually, but then met as a group to resolve
Porter and Steers's (1973) classic definition of met expecta- differences in ratings of job factors, so that a consensus descrip-
tions states that people compare their preentry expectations tion of organizational reality was obtained (Dean & Wanous,
and postentry perceptions. This requires a within-person com- 1982).
parison with measurements at two points in time. In our review, Using between-person discrepancy scores clearly violates the
however, we found a number of other types of definitions in- original definition of met expectations, because such measures
volving different comparisons with expectations and some- are insensitive to situations in which newcomers experience the
times involving one-shot measurement rather than longitudinal organization differently. However, future research using alter-
data. native formulations of between-person discrepancy scores may
One variation on Porter and Steers's (1973) version was to ask be able to determine the seriousness of this problem.
newcomers a single question after they had entered an organiza-
tion: "Think about what you expected to experience in this Work Environment Clarity Versus Ambiguity
organization and compare it to what you now experience, then
rate the direction and degree of the discrepancy between these The significance of work environment clarity versus ambigu-
two" (Wall & Payne, 1973). This approach was originally sug- ity for met expectations can be seen by asking this question:
gested by Wall and Payne as an alternative to using raw discrep- Under what conditions will a newcomer's expectations be dis-
ancy scores, which have a number of psychometric problems confirmed? Because newcomers learn from the insiders in the
(Cronbach & Furby, 1970). Although this method does avoid organization, the degree to which the messages they receive are
the use of discrepancy scores, it also precludes the gathering of clear and unambiguous will determine the degree of disconfir-
longitudinal data, relying instead on the accuracy of the re- mation. Environments that send clear messages to newcomers
spondent's memory. and in which there is relatively high consensus among the in-
A related issue in need of further investigation is the relative siders sending these messages have the potential to disconfirm
effect of expectations and perceptions on newcomer attitudes even the most strongly held expectations by newcomers. On the
and behavior. Even though the theory of met expectations im- other hand, environments in which newcomers receive ambigu-
plies that a discrepancy score using both expectations and per- ous or conflicting messages from insiders will allow new-
ceptions should be used to measure the concept, discrepancy comers to maintain their initial expectations (Colella, 1989;
scores have their problems, as noted previously. We were not Wanous & Colella, 1989). It is possible, then, that met expecta-
able to ascertain the possible effects of raw expectations, for tions theory applies only in the former type of environment,
example, because such data were not reported separately in but not in the latter. Much research on social cognition (e.g.,
most of the studies reviewed here. Future research should re- Fiske & Taylor, 1984; Markus & Zajonc, 1985; Taylor & Crocker,
 EFFECTS OF MET EXPECTATIONS 295

1981) shows that expectations can focus the attention of people Fiske, S. T, & Taylor, S. E. (1984). Social cognition. Reading, MA: Ad-
in such a way as to lead to their confirmation. The power of this dison-Wesley.
expectation effect has been ignored by met expectations re- Greenhaus, J. H., Seidel, C., & Marinis, M. (1983). The impact of ex-
searchers until recently (Colella, 1989), and it has yet to be pectations and values on job attitudes. Organizational Behavior and
Human Performance, 31, 394-417.
thoroughly researched.
Homer, S. O. (1979, September). A field experimental study of the affec-
tive, intentional and behavioral effects of organizational entry expecta-
References tions. Paper presented at the 87th Annual Convention of the Ameri-
can Psychological Association, New York.
Abbott, B. B., & Badia, P. (1986). Predictable versus unpredictable Hunter, J. E., & Hunter, R. F. (1984). Validity and utility of alternative
shock conditions and physiological measures of stress: Reply to predictors of job performance. Psychological Bulletin, 96, 72-98.
Arthur. Psychological Bulletin, 110, 384-387. Hunter, J. E., & Schmidt, F. L. (1990). Methods ofmeta-analysis. New-
Abbott, B. B., Schoen, L. S., & Badia, P. (1984). Predictable and unpre- bury Park, CA: Sage.
dictable shock: Behavioral measures of aversion and physiological laffaldano, M. T, & Muchinsky, P. M. (1985). Job satisfaction and job
measures of stress. Psychological Bulletin, 96, 45-71. performance: A meta-analysis. Psychological Bulletin, 67, 251-273.
Abelson, R. P., Aronson, E., McGuire, W J., Newcomb, T. M., Rosen- Johns, G. (1981). Difference score measures of organizational vari-
berg, M. J., & Tannenbaum, P. H. (Eds.). (1968). Theories of cognitive ables: A critique. Organizational Behavior and Human Performance,
consistency. Chicago: Rand McNally. 27, 443-463.
Adams, J. S. (1963). Toward an understanding of inequity. Journal of
Kemery, E. R., Mossholder, K. W, & Dunlap, W P. (1989). Meta-analy-
Abnormal and Social Psychology, 67, 422-436.
sis and moderator variables: A cautionary note. Journal of Applied
Arthur, A. Z. (1986). Stress of predictable and unpredictable shock.
Psychology, 74,168-170.
Psychological Bulletin, 100, 379-383.
Lawler, E. E. 111(1971). Pay and organizational effectiveness. New York:
Bass, A. R., & Ager, J. (1991). Correcting point-biserial correlations for
McGraw-Hill.
comparative analysis. Journal of Applied Psychology, 76, 595-598.
Locke, E. A. (1976). The nature and causes of job satisfaction. In M. D.
Caldwell, D. V, & O'Reilly, C. A. III. (1990). Measuring person-job fit
Dunnette (Ed.), Handbook of industrial and organizational psychol-
with a profile comparison process. Journal of Applied Psychology,
ogy (pp. 1297-1349). Chicago: Rand McNally.
75, 648-657.
March, J. G, & Simon, H. A. (1958). Organizations. New York: Wiley.
Callender, J. C., & Osburn, H. G. (1988). Unbiased estimation of sam-
Markus, H., & Zajonc, R. B. (1985). The cognitive perspective in social
pling variance of correlations. Journal of Applied Psychology, 73,
psychology. In G. Lindsey & E. Aronson (Eds.), The handbook of
312-315.
social psychology (Vol. 1,3rd ed., pp. 137- 230). New York: Random
Campbell, J. P., & Pritchard, R. (1976). Motivation theory in industrial
House.
and organizational psychology. In M. D. Dunnette (Ed.), Handbook
of industrial and organizational psychology (pp. 63-130). Chicago: McGuire, W J. (1985). Attitudes and attitude change. In G. Lindsey &
Rand McNally. E. Aronson (Eds.), The handbook of social psychology (Vol. 2,3rd ed.,
Colella, A. (1989). A new role for newcomer pre-entry expectations dur- pp. 233-344). New York: Random House.
ing organizational entry: Expectation effects on job perceptions. Un- Mobley, W H. (1982). Employee turnover: Causes, consequences, and
published doctoral dissertation, The Ohio State University, Colum- control. Reading, MA: Addison-Wesley.
bus, OH. Mowday, R. T. (1980, August). Unmet expectations about unmet expec-
Cook, R. D. (1977). Detection of influential observations in linear tations: Employee reactions to disconfirmed expectations during the
regression. Technometrics, 19, 15-18. early employment period. Paper presented at the 40th Annual Meet-
Cotton, J. L., & Tuttle, J. M. (1986). Employee turnover: A meta-analy- ing of the Academy of Management Association, Detroit.
sis and review with implications for research. Academy of Manage- Mowday, R. J., Steers, R. M., & Porter, L. W (1979). The measurement
ment Review, 11, 55-70. of organizational commitment. Journal of Vocational Behavior, 14,
Cronbach, L. J., & Furby, L. (1970). How should we measure "change" 224-247.
—or should we? Psychological Bulletin, 74, 68-80. Orr, J. M., Sackett, P. R., & Dubois, C. L. Z. (1991). Outlier detection
Cyert, R. M., & March, J. G. (1963). A behavioral theory of the firm. and treatment in I/O psychology: A survey of researcher beliefs and
Englewood Cliffs, NJ: Prentice-Hall. an empirical illustration. Personnel Psychology, 44, 473-486.
Dean, R. A. (1981). Reality shock, organizational commitment, and Pearlman, K., Schmidt, F. L., & Hunter, J. E. (1980). Validity generaliza-
behavior: A realistic job preview experiment (Doctoral dissertation, tion results for tests used to predict job proficiency and training
Michigan State University, 1981). Dissertation Abstracts Interna- success in clerical occupations. Journal of Applied Psychology, 65,
tional, 42, 12A, 5226. (University Microfilms, No. 82-12,381) 373-406.
Dean, R. A., & Wanous, J. P. (1982, August). Newcomers, insiders, and Petty, M. M., McGee, G. W, & Cavender, J. W (1984). A meta-analysis
old-timers: The effects of experience in an organization. Paper pre- of the relationship between individual job satisfaction and individ-
sented at the 42nd Annual Meeting of the Academy of Management ual performance. Academy of Management Review, 9, 712-721.
Association, New York. Porter, L. W, Lawler, E. E. Ill, & Hackman, J. R. (1975). Behavior in
Edwards, J. R., & Cooper, C. L. (1990). The person-environment fit organizations. New York: McGraw-Hill.
approach to stress: Recurring problems and some suggested solu- Porter, L. W, & Steers, R. M. (1973). Organizational, work, and per-
tions. Journal of Organizational Behavior, 11, 293-308. sonal factors in employee turnover and absenteeism. Psychological
Feldman, D. C. (1976). A contingency theory of socialization. Adminis- Bulletin, 80, 151-176.
trative Science Quarterly, 21, 433-452. Premack, S. L., & Wanous, J. P. (1985). A meta-analysis of realistic job
Fiedler, F. E. (1967). A theory of leadership effectiveness. New York: preview experiments. Journal of Applied Psychology, 70, 706-719.
McGraw-Hill. Rasmussen, J. L., & Loher, B. T. (1988). Appropriate critical percent-
Fiedler, F. E., & Garcia, J. E. (1987). New approaches to effective leader- ages for the Schmidt and Hunter meta-analysis procedure: Compara-
ship: Cognitive resources and organizational performance. New \brk: tive evaluation of Type I error rate and power. Journal of Applied
Wiley. Psychology, 73, 683-687.
296 WANOUS, POLAND, PREMACK, AND DAVIS

Rice, R. W, McFarlin, D. B., & Bennett, D. E. (1989). Standards of Vroom, V H. (1964). Work and motivation. New York: Wiley.
comparison and job satisfaction. Journal of Applied Psychology, 74, Vroom, V H., & Yetton, P. W (1973). Leadership and decision making.
591-598. Pittsburgh, PA: University of Pittsburgh Press.
Rosenthal, R. (1979). The "file drawer problem" and tolerance for null Wall, T. D., & Payne, R. (1973). Are deficiency scores deficient? Journal
results. Psychological Bulletin, 86, 638-641. of Applied Psychology. 58, 322-326.
Schein, E. A. (1978). Career dynamics: Matching individual and organi- Wanous, J. P. (1977). Organizational entry: Newcomers moving from
zational needs. Reading, MA: Addison-Wesley. outside to inside. Psychological Bulletin, 84, 601-618.
Schmitt, N., Gooding, R. Z., Noe, R. D, & Kirsch, M. (1984). Meta-an- Wanous, J. P. (1980). Organizational entry. Reading, MA: Addison-
alysis of validity studies published between 1964 and 1982 and the Wesley.
investigation of study characteristics. Personnel Psychology, 37,407- Wanous, J. P. (1992). Organizational entry: Recruitment, selection, orien-
422. tation, and socialization. Reading, MA: Addison-Wesley.
Schneider, B. (1975). Organizational climate: Individual preferences Wanous, J. P., & Colella, A. (1989). Organizational entry research:
and organizational realities revisited. Journal of Applied Psychology, Current status and future directions. In G. Ferris & K. Rowland
60, 459-465. (Eds.), Research in personnel and human resource management (Vol.
Schneider, B., & Schmitt, N. (1986). Staffing organizations (2nd ed.). 7, pp. 59-120). Greenwich, CT: JAI Press.
Glenview, 1L: Scott, Foresman. Wanous, J. P., Keon, T, & Latack, J. C. (1983). Expectancy theory and
Spector, P. E. & Levine, E. L. (1987). Meta-analysis for integrating study occupational/organizational choices: A review and test. Organiza-
outcomes: A Monte Carlo study of its susceptibility to Type I and
tional Behavior and Human Performance, 32, 66-86.
Type II errors. Journal of Applied Psychology, 72, 3-9.
Wanous, J. P., &Lawler, E. E. III. (1972). Measurement and meaning of
Stevens, J. P. (1984). Outliers and influential data points in regression
job satisfaction. Journal of Applied Psychology, 56, 95-105.
analysis. Psychological Bulletin, 95, 334-344.
Taylor, S. E. & Crocker, J. (1981). Schematic bases of social information Wanous, J. P., Sullivan, S. E., & Malinak, J. (1989). The role of judgment
processing. In E. T. Higgins, C. P. Herman, & M. P. Zanna (Eds.), calls in meta-analysis. Journal of Applied Psychology, 74, 259-264.
Social cognition: The Ontario Symposium on Personality and Social Weitz, J. (1956). Job expectancy and survival. Journal of Applied Psy-
Psychology. Hillsdale, NJ: Erlbaum. chology, 40, 245-247.
Tukey, J. W(1977). Exploratory data analysis. Reading, MA: Addison- Williams, C. R. (1990). Deciding when, how, and if to correct turnover
Wesley. correlations. Journal of Applied Psychology, 75, 732-737.
Van Maanen, J. (1976). Breaking in: Socialization to work. In R. Dubin Youngberg, C. F (1963). An experimental study of job satisfaction and
(Ed.), Handbook of work, organization, and society (pp. 67-130). turnover in relation to job expectations and self expectations. Dis-
Chicago: Rand McNally. sertation Abstracts. 25, 656. (University Microfilms No. 64-6516)
 EFFECTS OF MET EXPECTATIONS 297

Appendix

Studies Used in the Meta-AnalysisAl

1. Arnold, H. J., & Feldman, D. C. (1982). A multivariate analysis of and women. Journal of Applied Psychology, 72, 393-400. (O, NON,
the determinants of job turnover. Journal of Applied Psychology, 67, PUB)
350-360. (O, NON, PUB) 19. Lee, T. W., & Mowday, R. T. (1987). Voluntarily leaving an organi-
2. Avner, B. K., Guastello, S. J., & Aderman, M. (1982). The effect of zation: An empirical investigation of Steers and Mowday's model of
a realistic job preview on expectancy and voluntary versus involun- turnover. Academy of Management Journal, 30, 721-743. (PS,
tary turnover. Journal of Psychology, 111, 101-107. (O, EXP, PUB) NON, PUB)
3. Blau, G. (1988). An investigation of the apprenticeship organiza- 20. Michaels, C. E., & Spector, P. E. (1982). Causes of employee turn-
tional socialization strategy. Journal of Vocational Behavior, 32, over: A test of the Mobley, GrifFeth, Hand, and Meglino model.
176-195. (PS, NON, PUB) Journal of Applied Psychology, 67, 53-59. (O, NON, PUB)
4. Colella, A. J. (1989). A role for newcomer pre-entry expectations 21. Pierce, J. L., & Dunham, R. B. (1987). Organizational commit-
during organizational entry: Expectation effects on job perceptions. ment: Pre-employment propensity and initial work experiences.
Unpublished doctoral dissertation, The Ohio State University, Co- Journal of Management, 13, 163-178. (PS, NON, PUB)
lumbus. (PS, NON, UNP) 22. Popovich, P. M. (1986, August). Characteristics of the job satisfac-
5. Dean, R. A. (1981). Reality shock, organizational commitment, tion of new employees. Paper presented at the 46th Annual Meeting
and behavior: A realistic job preview experiment. Unpublished doc- of the Academy of Management, Chicago, IL. (PS, NON, UNP)
toral dissertation, Michigan State University, East Lansing. (PS, (This study was considered an outlier after computation of the initial
NON, UNP) meta-analysis.)
6. Dean, R. A., & Wanous, J. P. (1984). Effects of realistic job pre- 23. Pulakos, E. D., & Schmitt, N. (1983). A longitudinal study of a
views on hiring bank tellers. Journal of Applied Psychology, 69, 61 - valence model approach for the prediction of job satisfaction of new
68. (PS, EXP, PUB) employees. Journal of Applied Psychology, 68, 307-312. (PS, NON,
7. Dugoni, B. L., & Ilgen, D. R. (1981). Realistic job previews and the PUB)
adjustment of new employees. Academy of Management Journal, 24. Reichers, A. E., & Kim, J. S. (1987). A longitudinal investigation of
24, 579-591. (O, EXP, PUB) an interactionist perspective on newcomer socialization. Unpub-
8. Farkas, A. J., & Tetrick, L. E. (1989). A three-wave longitudinal lished manuscript, Ohio State University, College of Business, Co-
analysis of the causal ordering of satisfaction and commitment on lumbus. (O, NON, UNP)
turnover decisions. Journal of Applied Psychology, 74, 855-868. (PS, 25. Reilly, R. R., Brown, B., Blood, M. R., & Malatesta, C. Z. (1981).
NON, PUB) The effects of realistic previews: A study and discussion of the litera-
9. Federico, S. M., Federico, P., & Lundquist, G. W. (1976). Predict- ture. Personnel Psychology, 34, 823-834. (O, EXP, PUB)
ing women's turnover as a function of extent of met salary expecta- 26. Schneider, B. (1975). Organizational climate: Individual prefer-
tions and biodemographic data. Personnel Psychology, 29, 559-566. ences and organizational realities revisited. Journal of Applied Psy-
(O, NON, PUB) chology, 60, 459-465. (O, NON, PUB)
10. Feldman, D. C. (1976). A contingency theory of socialization. Ad- 27. Stumpf, S. A., & Hartman, K. (1984). Individual exploration to
ministrative Science Quarterly, 21, 433-452. (O, NON, PUB) organizational commitment or withdrawal. Academy of Manage-
11. Fisher, C. D. (1985). Social support and adjustment to work: A ment Journal, 27, 308-329. (O, NON, PUB)
longitudinal study. Journal of Management, 11, 39-53. (PS, NON, 28. Suszko, M. K., & Breaugh, J. A. (1984). The effects of realistic job
PUB) previews on applicant self-selection and employee turnover, satisfac-
12. Greenhaus, J. H., Seidel, C., & Marinis, M. (1983). The impact of tion, and coping ability. Journal of Management, 12, 513-523. (PS,
expectations and values on job attitudes. Organizational Behavior EXP, PUB)
and Human Performance, 31, 394-417. (PS, NON, PUB) 29. Vanderberg, R. J., & Scarpello, V. (1990). The matching model:
13. Hoiberg, A. & Berry, N. H. (1978). Expectations and perceptions An examination of the processes underlying realistic job previews.
of Navy life. Organizational Behavior and Human Performance, 21, Journal of Applied Psychology, 75, 60-67. (PS, NON, PUB)
130-145. (O, NON, PUB) 30. Youngberg, C. F. (1963). An experimental study of job satisfaction
14. Horner, S. D. (1979, September). A field experimental study of the and turnover in relation to job expectations and self expectations.
affective, intentional, and behavioral effects of organizational entry Dissertation Abstracts, 25, 656. (University Microfilms No. 64-
expectations. Paper presented at the 87th Annual Convention of the 6516). (PS, EXP, UNP) (This study was considered an outlier after
American Pychological Association, New York. (PS, EXP, UNP) computation of the initial meta-analysis.)
15. Ilgen, D. R. (1971). Satisfaction with performance as a function of 31. Youngblood, S. A., Mobley, W. H., & Meglino, B. M. (1983). A
the initial level of expected performance and the deviation from longitudinal analysis of the turnover process. Journal of Applied Psy-
expectations. Organizational Behavior and Human Performance, 6, chology, 68, 507-516. (O, NON, PUB)
345-361. (PS, EXP, PUB) (This study was considered an outlier
after computation of the initial meta-analysis.)
16. Katerberg, R., & Blau, G. (1980). A study of met expectations in Al
Each study is coded three ways: (a) whether it conforms to Porter and
organizational assimilation. Unpublished manuscript, University of Steers's (1973) definition (PS), or whether it uses some other (O) defini-
Cincinnati, Department of Management, Cincinnati, OH. (PS, tion; (b) whether it is experimental (EXP) or nonexperimental (NON);
NON, UNP) and (c) whether it is published (PUB) or unpublished (UNP).
17. Katzell, M. E. (1968). Expectations and dropouts in schools of
nursing. Journal of Applied Psychology, 72, 393-400. (PS, NON,
PUB) Received May 28, 1991
18. Latack, J. C, Josephs, S. L., Roach, B. L., & Levine, M. C. (1987). Revision received December 17, 1991
Carpenter apprentices: Comparison of career transitions for men Accepted December 19, 1991 •