Research Theory, Design, and Methods
Walden University
Purpose Statement Checklist
Use the following criteria to evaluate an author’s purpose statement.
Look for indications of the following:
•
Does the statement begin with signaling words?
•
Does the statement identify the research approach (quantitative,
qualitative, or mixed)?
•
Does the statement clearly state the intent of the study?
•
Does the statement mention the participants?
•
Does the statement mention the research site?
•
Is the statement framed in a way that is consistent with the identified
problem?
If the study is qualitative, does the purpose statement do as follows?
•
Focus on a single phenomenon
•
Use an action verb to convey how learning will take place
•
Use neutral, nondirectional language
•
Provide a general definition of the central phenomenon
If the study is quantitative, does the purpose statement do as follows?
•
Identify the variables under study
•
Provide a general definition of each key variable
•
Use words that connect the variables
•
Identify a theory
If the study is mixed methods, does the purpose statement do as follows?
•
Discuss the reason(s) for mixing both quantitative and qualitative data
•
Include the characteristics of a good qualitative purpose statement (as
listed above)
© 2016 Laureate Education, Inc.
Page 1 of 2
Research Theory, Design, and Methods
Walden University
•
Include the characteristics of a good quantitative purpose statement (as
listed above)
•
Indicate the specific method of collecting both quantitative and qualitative
data
© 2016 Laureate Education, Inc.
Page 2 of 2
Article
Public Service Motivation and
Institutional-Occupational
Motivations Among
Undergraduate Students and
ROTC Cadets
Public Personnel Management
2014, Vol. 43(4) 442-458
© The Author(s) 2 0 14
Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1 177/0091026014530270
ppm.sagepub.com
®SAGE
Katherine M. Ngaruiya1, Anne-Lise Knox V elez1,
Richard M. Clerkin’, and Jami Kathleen Taylor 2
A b strac t
Given the current fiscal climate, budgetary pressures may have important implications
for recruitment and retention of military personnel. In response to this issue, we
join two literatures to study motivational differences in undergraduate college
students and Reserve Officer Training Corps (ROTC) cadets: Moskos’ Institutional
and Occupational (l-O) enlistment motivation model and Kim et al.’s revised Public
Service Motivation (PSM) scale. We survey ROTC cadets and undergraduates at a
mid-size public university and find that PSM is higher for ROTC cadets than regular
undergraduates. We also find that for ROTC cadets, the institutional motivators for
enlistment correlate positively with the rational, normative, and affective dimensions
of PSM. In addition, we find increases in the Occupational motivator and the
compassion PSM dimension reduce the likelihood of being an ROTC cadet whereas
the Institutional motivator and the self-sacrifice PSM dimension are positively related
with being an ROTC cadet.
Keywords
public service motivation, motivation theory, workplace attitudes and behaviors,
public management
‘N o rth Carolina State University, Raleigh, NC, USA
2University of Toledo, Toledo, OH, USA
Corresponding Author:
Katherine M. Ngaruiya, Department of Public Administration, School o f Public and International Affairs,
North Carolina State University, Caldwell Hall, Campus Box 8 102, Raleigh, NC 27695, USA.
Email: kmngaruiya@ncsu.edu
Ngaruiya et al.
443
In tr o d u c tio n
Despite recent conflicts in Afghanistan and Iraq, the U.S. military faces future funding
cuts. Due to unsustainable budget deficits, President Obama’s administration recently
proposed significant budget cuts to the Department of Defense (Dreazen, 2012).
Budget cuts will likely affect salaries for military personnel and possible reductions in
extrinsic motivators might pose barriers to recruitment, performance, and retention of
personnel.
This study investigates a potential way to mitigate these looming problems by con­
necting two streams of research that address the employment motivations of public
sector workers. The first body of work is public service motivation (PSM). It has been
used to assess the intrinsic motivations of individuals who pursue careers in the public
sector. The second body of literature is derived from Moskos’ (1977, 1986)
Institutional-Occupational (I-O) model. This literature has been used to explore the
enlistment and retention motivations of a subset of public workers—military person­
nel. By joining these previously separate streams of research, it is hoped that dialogue
between researchers in these two fields will provide substantive insights on career
selection, performance, and the effect of reward systems.
In this article, we first review the development of Moskos’ 1-0 model and PSM.
We then test whether there is an overlap between PSM and the Institutional Dimension
of Moskos’ model. We do so by administering an instrument consisting of PSM (Kim
et al., 2012) and 1-0 questions to undergraduate students and Reserve Officer Training
Corps (ROTC) cadets at a mid-sized Midwestern university. The ROTC is a program
offered at more than 1,000 colleges across the United States that prepares young
adults to become officers in the U.S. military. “In exchange for a paid college educa­
tion and post-college career, cadets commit to serve in the Military after gradua­
tion. .. [and] each Service branch has its own take on ROTC” (Today’s Military, n.d.).
Given the literature on military service and sacrifice, we expect that PSM and
Moskos’ Institutional motivation to be more highly correlated among ROTC cadets
than among other undergraduates. We then perform a logistic regression analysis to
test whether we can predict the likelihood of a respondent being an ROTC cadet
based on models that include the PSM and 1-0 factors. Although this is an explor­
atory research on a complex issue, we believe that the conclusions derived from this
study are a first step in providing new insights on creating efficient and effective
recruitment mechanisms.
E n l i s t m e n t a n d R e t e n t i o n M o t i v a t i o n s in t h e M i l i t a r y
A person’s decision to join the military is complex and is commonly motivated by a
number of factors (Ginexi, Miller, & Tarver, 1994). The combined influences of
“social, personal and organizational factors” affect military enlistment (Mehay, 1990,
p. 364). These tangible (extrinsic) and intangible (intrinsic) motivators impact propen­
sity to serve in the military (Griffith, 2008). Tangible motivators include salary, ben­
efits, enlistment bonus, and money for college (Woodruff, Kelty, & Segal, 2006).
Intangible motivators include desire for self-improvement, desire to serve others,
444
Public Personnel M anagem ent 43(4)
aspiration to serve one’s country, and becoming disciplined and confident (Griffith,
2008; Lawrence & Legree, 1996). Researchers exploring these motivators have often
categorized these motives based on Moskos’ 1-0 model (e.g., Griffith, 2008; Woodruff
et al., 2006). While the 1-0 model is organizational in its level of analysis, its insights
have been useful in analyzing the diverse individual level motivations about military
service. The Institutional military is one in which soldiers serve in response to a call to
duty and honor (Moskos, 1977; Woodruff et ah, 2006). In contrast to the Institutional
military, Moskos identifies the Occupational military as one in which the free market
dominates military service and its members (Moskos, 1977). Thus, a desire like patri­
otic duty could be an institutional motive while job training is an occupational-related
influence. In general, intrinsic motivation is associated with Moskos’ su m m a tio n of
“Institutional orientation” whereas the “Occupational orientation” relates to educa­
tional benefits, earning money, and receiving money as incentives to enlist.
Moskos’ 1-0 model has been used to study recruitment and retention of military
personnel by a range of scholars. For instance, Eighmey (2006) identified seven
themes related to youth enlistment: benefits, fidelity (desire to serve community and
duty to country), dignity (pride in work, working in environment free of discrimina­
tion), risk, family (approval/respect from family and friends), challenge, and adven­
ture. These themes were then transposed onto the Moskos’ 1-0 model; youth in the
study distinguished between tangible and intangible aspects of self (Occupational) and
other oriented goals (Institutional; Eighmey, 2006). Griffith (2008) notes that
Institutionally motivated reservists are more likely to reenlist than their Occupationally
motivated peers. Woodruff et al. (2006) found that Institutional motivators are cen­
trally important to determining enlistment propensity in high school seniors.
Segal and Segal (2004) found additional enlistment factors for student populations.
Students with fewer educational prospects after high school have a higher propensity
to enlist. Conversely, individuals who performed well in high school and are children
of college-educated parents are less likely to enlist. Although the Occupational vari­
ables (extrinsic values) have historically been used as the bait to draw in new military
recruits in public relations and advertising campaigns, scholars have noted that more
research should be devoted to studying the Institutional considerations (internal val­
ues) that lead to enlistment decisions (Eighmey, 2006). To add some depth to the dis­
cussion on internal values that effect enlistment decisions, we tie Moskos’ ideas to
PSM. Military personnel are public sector workers who have largely been ignored in
PSM research.1 The following section discusses the connection between PSM and the
1-0 model as it reviews the literature on PSM.
PSM
Vandenabeele (2007) defines PSM as “belief, values, and attitudes that go beyond
self-interest or organizational interest, that concern the interest of a larger political
entity and that motivate individuals to act accordingly whenever appropriate” (p. 547).
It is commonly agreed that PSM is a multi-dimensional construct (Kim et al., 2012;
Perry & Wise, 1990). As originally posited by Perry and Wise (1990), PSM consisted
of three classes of motivators that spur public service: rational (pursuit of policy
N g a r u iy a
et
al.
445
objectives), affective (emotional attachment to a group or community), and normative
(a sense of duty or desire to “give back”). It is the normative aspect of PSM that may
provide the closest connection to the institutional motivators related to military
service.
In Perry and Wise’s (1990) seminal work, PSM is not bounded by “locus of employ­
ment . . . but may be understood as an individual’s predisposition” (p. 368) to respond
to motives to serve in primarily public institutions and organizations. PSM has been
applied to motivational differences between public and private sector workers (e.g.,
Brewer & Selden, 1998; Clerkin & Coggburn, 2012; Houston, 2005; Perry & Wise,
1990). It has also been extended to address the motivations of nonprofit workers
(Perry, 2000) and volunteers (Clerkin, Paynter, & Taylor, 2009; Coursey, Perry,
Brudney, & Littlepage, 2008; Houston, 2005; Perry, Brudney, Coursey, & Littlepage,
2008). People who work or volunteer in government agencies and nonprofit organiza­
tions often have higher levels of PSM than do private sector employees.
Although mediated by contextual factors such as whether the person fits the orga­
nization (Brewer, 2008; Bright, 2007; Wright & Pandey, 2008), research supports the
notion that PSM is positively related to public/nonprofit sector job satisfaction and
performance (Naff & Crum, 1999; Perry, Hondeghem, & Wise, 2010). It is also nega­
tively associated with employee attrition (e.g., Crewson, 1997; N aff & Crum, 1999;
Steijn, 2008).
PSM has also been used to investigate incentive preferences in public sector work­
ers. Some scholars have found an inverse relationship between PSM and valuation of
monetary rewards (Bright, 2007, 2009; Karl & Sutton, 1998). However, Crewson
(1997) notes a more nuanced relationship by finding that although high pay is impor­
tant for those in public service, other non-financial extrinsic incentives are less impor­
tant for public sector workers than private sector employees. This is consistent with
Wise’s (2004) finding that public sector workers often have multiple motives to seek
employment in the public sphere. This review of employment incentives provides
another connection between the two literatures given that the Occupational dimension
of the 1-0 model addresses extrinsic reward preferences.
Recently, Kim and Vandenabeele (2010) propose clarifications to Perry’s (1996)
original PSM measurement model to improve its psychometric reliability and validity.
Kim and Vandenabeele’s (2010) measurement of PSM contains four dimensions as
follows: attraction to public participation (APP), commitment to public values (CPV),
compassion (COM), and self-sacrifice (SS). APP focuses on the extent to which indi­
viduals want to participate in the policy process or other activities that contribute to the
larger society. CPV measures the extent to which an individual’s interest in public
service is driven by an interest in public values like equity and ethics. The COM
dimension is based on individual concern for the needs of specific individuals or
groups and SS is related to altruistic and pro-social origins of PSM. Kim (2011) con­
cluded the true essence of PSM is the willingness to provide service to others without
receiving tangible personal rewards in return.
Adapting Perry’s (1996) 24 question scale, Kim (2011) later surveyed more than
2,000 Korean firefighters to confirm the PSM construct as having four dimensions
446
Public Personnel M anagem ent 43(4)
(APM, CPV, COM, SS) that are based on instrumental, values-based and affective
motifs as well as being altruistic and having pro-social origins. A recent article by Kim
and a number of colleagues (2012) recommended changes to the PSM measurement
tool. After conducting a cross-national study, they proposed a revised 20-item PSM
survey instrument renaming the APP dimension to attraction to public service (APS;
Appendix A). Kim et al.’s (2012) revised tool is used as the basis for the surveys and
results presented in this article.
Based on the literature on PSM and Moskos’ 1-0 model, holding all else constant,
we hypothesize the following:
Hypothesis 1: Within the ROTC sample, the Institutional motivators for enlistment
will correlate positively with the normative dimension of PSM (CPV).
Hypothesis 2: CPV will be higher for ROTC cadets than for undergraduate
students.
Hypothesis 3: The higher an individual’s CPV, the greater the likelihood that they
will be an ROTC cadet.
D a ta a n d M e th o d
To investigate our hypotheses, we surveyed undergraduate students and ROTC cadets
enrolled at a mid-sized public university in the Midwest. Our survey instrument com­
bined the 20-item PSM instrument developed by Kim et al. (2012) with Likert-type
scale 1-0 questions adopted from work by Woodruff et al. (2006). While the PSM
questions remained the same on surveys administered to each sample, the Moskos’ 1-0
questions were modified to account for the ROTC/Non ROTC status of our respon­
dents (Appendices B and C). In addition, the instrument contained demographic ques­
tions that captured the respondent’s race, gender, and religiosity. Data were collected
via online administration of our instrument to 290 undergraduate and 104 ROTC stu­
dents during the fall semester of 2011. The undergraduate portion of the sample was
drawn from a research pool of students enrolled in introductory American government
courses. These courses are used to fill a “general education” requirement at the univer­
sity and thus they provide a sample of students that is reasonably reflective of the
university’s undergraduate population.2 ROTC students were accessed with permis­
sion of their command structure.
Pairwise partial correlations were performed to assess the relationships between
the PSM dimensions and Moskos’ 1-0 motivators (Hypothesis 1) in the sample of
ROTC cadets. An independent-samples test was performed to assess the differences
in means between the ROTC respondents and the undergraduate students (Hypothesis
2). Logistic regression was used to predict the likelihood of being an ROTC cadet,
using the PSM dimensions, Moskos’ 1-0 dimensions, and other demographic vari­
ables in the model (Hypothesis 3). The dependent variable is a dichotomous indicator
of whether a student is in the ROTC (0 for not ROTC and 1 for ROTC). The follow­
ing section describes construction of the indices that comprise our independent
variables.
Ngaruiya et al.
447
T ab le 1. Control Variables Descriptive Statistics.
White
Religious
Female
n
Overall
Undergraduate
ROTC
71%
17%
42%
384
67%
18%
47%
290
85%
15%
27%
94
Note. R O TC = Reserve O fficer Training Corps.
Inde pe n de n t Variables o f Interest
Cronbach’s alpha models internal consistency of an index based on average correla­
tion among items (Garson, 2012b). To assess the reliability of the various scales used
in this study, we calculated the Cronbach’s alpha coefficient for the four Kim et al.
(2012) PSM dimensions. Appendix A provides the indicators for each dimension. Per
Garson (2012b), a lenient cutoff of .60 is common in exploratory research; alpha
should be at least .70 or higher to retain an item in an “adequate” scale, and many
researchers require a cutoff of .80 for a “good scale.” The reliability scores for the four
PSM dimensions are .773 for APS, .672 for CPV, .731 for COM, and .781 for SS.
Thus, APS, COM, and SS are considered adequate, but CPV is acceptable only at the
exploratory level. PSM scores for each respondent were calculated based on their
average scores for APS, CPV, COM, and SS dimensions. The reliability scores for
Moskos’ 1-0 Dimensions are .821 for Institutional, indicating a high level of internal
reliability and .721 for Occupational, indicating an adequate scale. Appendices B and
C provide the indicators for these constructs.
C ontrol Variables
We include three control variables in our regression models, whether a respondent is
White, whether they are female, and their level of religiosity to control for potential
demographic differences between ROTC cadets and the undergraduate respondents.
The descriptive statistics for the ROTC and undergraduate respondents are reported
below in Table 1. ROTC cadets are more likely to be White, less religious, and less
likely to be female than the undergraduate students in our sample.
Analysis and Results
ROTC C adet Pairwise C orrelations Analyses
In our first hypothesis, we posit that the Moskos’ Institutional motivators for enlist­
ment will correlate positively with the normative dimension of PSM (CPV) within the
ROTC sample. Pairwise correlations were performed on the ROTC sample to assess
the relationships between the PSM Dimensions and Moskos’ 1-0 dimensions (see
448
Public Personnel M anagem ent 43(4)
Partial Correlation Between Public Service Motivation Dimensions and Moskos’
Motivations ROTC Cadets.
T a b le 2.
APS
APS
CPV
SS
COM
Institutional
Occupational
1
.580**
.625**
.619**
.281*
-.222*
CPV
1
^92**
429**
.220*
-.116
SS
1
.506**
.308*
-.257*
COM
Institutional
Occupational
1
.086
.013
1
-.101
1
Note, n = 94. R O TC = Reserve O fficer Training Corps; APS = attraction to public service; CPV =
co m m itm ent to public values; C O M = compassion; SS = self-sacrifice.
*p < .05. **p < .001, tw o-tailed test.
Table 2).3 Within the ROTC sample, a small but statistically significant degree of cor­
relation was found between the CPV dimension of PSM and Moskos’ Institutional
dimension. Thus, we find moderate support for Hypothesis 1.
Two other PSM dimensions also have statistically significant correlations with
Moskos’ Institutional motivation. APS and SS are positively correlated with the
Institutional dimension. These additional positive correlations between the PSM
dimensions and Moskos’ Institutional motivation suggest that Institutional motivation
to join the military may actually be multi-dimensional. Examining motivations to join
the military through the dimensions of PSM might give us a more nuanced under­
standing of Institutional motivation to join the military.
The correlations between APS, and SS and the Occupational dimension were nega­
tive and statistically significant. The higher an ROTC cadets’ APS and SS, the lower
their level of Occupational motivation to join the ROTC. Interestingly, Moskos’ 1-0
motivations are negatively (but not statistically significantly) correlated among the
ROTC respondents.
Difference o f M eans Tests
To test Hypotheses 2, an independent-samples t test is performed to assess the differ­
ences in means of the CPV dimension of PSM between the ROTC respondents and the
undergraduate students (see Table 3). We find that the CPV means in the two groups
have statistically significant differences. As expected, ROTC cadets have a higher CPV
mean (4.45) than do other undergraduates (4.29). This provides evidence in support of
Hypothesis 2. We also find statistically significant differences between ROTC cadets
and undergraduates in two other PSM dimensions, APS and SS. In both cases, ROTC
cadets have higher levels of PSM than do undergraduates. Although our data help us
establish this difference, we are unable to know whether this difference existed before
enrolling in college, or if there is something in ROTC training that helps to develop and
increase PSM levels in cadets that are not manifest in other undergraduates.
449
Ngaruiya et at.
Table 3. D ifferences o f Means Between U ndergraduates and R O T C Cadets.
D ifferences in means
M
n
SD
APS
U ndergraduate
ROTC
290
94
4.1834*
0.60503
4.3915
0.53136
CPV
U ndergraduate
ROTC
4.2886*
0.50577
94
4.4496
0.47026
290
4.0786
0.65562
94
4.0064
0.60851
3.5578**
0.82591
4.1711
0.64554
290
COM
U ndergraduate
ROTC
SS
U ndergraduate
ROTC
290
94
In stitu tio n a l dim ension
U ndergraduate
ROTC
290
94
3.4853**
1.17500
4.5691
0.53557
O ccupational dim ension
U ndergraduate
ROTC
290
94
3.2230**
0.98039
2.6454
0.83693
N o te , n = 94. R O TC = Reserve O fficer Training Corps; APS = attraction to public service; CPV =
com m itm ent to public values; C O M = compassion; SS = self-sacrifice.
* p < .05. * * p < .0 0 1 , tw o-tailed test.
We also checked for differences in Occupational and Institutional motivations
between ROTC cadets and undergraduates. The Institutional motivators are signifi­
cantly higher in ROTC respondents, whereas Occupational motivators are signifi­
cantly higher in undergraduates. This finding potentially indicates that if the ROTC
wanted to increase the number of cadets from the undergraduate population on cam­
pus, focusing on Occupational motivators would be more effective than Institutional
or PSM motivators.
Logistic Regressions
To assess Hypothesis 3, four logistic regressions were utilized to predict ROTC mem­
bership in our combined sample of undergraduates and ROTC cadets. The dependent
variable in each model is a dichotomous indicator of whether a respondent is a ROTC
cadet (0 = not ROTC, 1 = ROTC). In each model, we include control variables and
introduce different combinations of our variables of interest from Moskos’ 1-0 and
Kim et al.’s (2012) PSM measurements. In Model 1, we include the 1-0 dimensions
from Moskos. In Model 2, we include the four PSM dimensions. Given the positive
correlation between Moskos’ Institutional dimension and most of the PSM dimen­
sions, in Model 3, we include the four PSM dimensions and the Occupational
450
Public Personnel M anagem ent 43(4)
T a b le 4. Logistic Regression Analysis to Predict ROTC Cadets.
Model 1
Odds
ratio
Model 2
z
Odds
ratio
Model 3
z
Odds
ratio
APS
1.054
0.13
1.112
CPV
1.387
0.88
1.386
COM
SS
Occupational
Institutional
W h ite
Model 4
z
Odds
ratio
0.27 0.545
0.86 0.828
z
-1.31
-0.41
0.277*** -4.06 0.323***-3.52 0.525t -1.79
5.56 4.629*** 5.12 3 507* * * 3.87
0.265*** 7.69
0 .586***-3.70 0.290*** -5.95
5.656*** -6.79
5 .104*** 6.76
1.154
0.37 2.790** 2.96 2.847** 2.94 1.287
0.60
5.176***
Religious
0.708
Female
0.613
-0.78 0.866
-0.37 0.774
-1.46 0.605
-1.68 0.613
Constant
0.017*** -4.1 1 0.010***-3.41 0.033*
% C orre ctly classified
85.26
77.63
81.32
Proportional reduction e rro r 40.43
9.57
24.47
(%)
-0.65 0.769
-1.60 0.709
-2.44 0.053+
-0.56
-0.95
-1.80
86.05
43.62
Note, n –
380. R O TC – Reserve O ffice r Training Corps; APS = attraction to public service; CPV =
com m itm ent t o public values; C O M = compassion; SS = self-sacrifice.
t p < . l . * p < .05. **p < . 0 1. ***p < .0 0 1, all tw o-tailed tests.
dimension to see if the PSM dimensions have similar predictive ability to Moskos’
Institutional dimension. Finally, in Model 4, we include all six dimensions of interest
to see if the PSM dimensions have direct effects on the likelihood of a respondent
being in the ROTC while controlling for Moskos’ Institutional dimension. Control
variables in all models include gender (whether the respondent is female), race
(whether the respondent is White), and religious attendance of respondents (dichoto­
mized into once a week or more frequent attendance of religious services over the past
year vs. less frequent or no attendance). All regression equations showed goodness of
fit through a significant Omnibus test. Table 4 provides the results of the four logistic
regressions. We turn to reviewing the results next.
Model 1 includes the 1-0 dimensions and the three control variables. Only the
1-0 factors were found to be significant predictors of ROTC service. For this model,
there was a proportional reduction in error of 40.43% compared with chance
(Garson, 2012a).
Model 2, which includes the four PSM dimensions (APS, CPV, COM, and SS) and
the three control variables, finds that COM and SS are statistically significant, although
in opposite directions. As SS increases, the likelihood of being an ROTC cadet
increases. However, as COM increases, the likelihood of being an ROTC cadet
decreases. In Model 2, the hypothesized relationship between CPV and ROTC service
is not supported. We find that CPV is positively related to the likelihood of being an
ROTC cadet but is not statistically significant. Although CPV is not statistically
N g a ru iya
et al.
451
significant, it performs in the expected direction in relation to ROTC service. Only one
control variable in Model 2 is statistically significant, that is, being White increases
the likelihood of being an ROTC cadet. The proportional reduction in error for this
model is only a 9.57%, indicating that PSM, by itself, is not a strong predictor of
whether a respondent is in the ROTC.
In Model 3, we add Moskos’ Occupational dimension back into the model with the
four PSM dimensions. The significance and tendency of the PSM variables do not
change between Models 2 and 3; neither APS nor CPV is statistically significant, SS
is positively related to the likelihood of being an ROTC cadet and COM is negatively
related to being an ROTC cadet. Increases in the Occupational dimension decrease the
likelihood of being an ROTC cadet. Finally, being White increases the likelihood of
being an ROTC cadet. For this model, the proportional reduction in error is 24.47%
compared with chance.
In our final model, Model 4, in addition to the control variables, we regress all of
the PSM and Moskos’ dimensions on being an ROTC cadet. Both of Moskos’
dimensions, Institutional (positively) and Occupational (negatively) are signifi­
cantly related to the likelihood of being an ROTC cadet. For the PSM dimensions,
CPV and APS are still not statistically significantly related to the likelihood of being
an ROTC cadet and SS is still statistically significant and positively related to being
an ROTC cadet. Increases in the COM dimension still reduce the likelihood of being
an ROTC cadet. However, with a z score o f -1.79 it is only statistically significant
at the p = .1 level for a two-tailed test. Combining both the Moskos and PSM dimen­
sions to predict the likelihood of being an ROTC cadet results in a proportional
reduction in error is 43.62%.
Overall, our data do not support Hypothesis 3, in which we expected a positive
relationship between CPV and ROTC service. Instead, we find that CPV is not related
to this likelihood at all. In addition, the degree of explanation or reduction in error,
when using PSM alone or with the Occupational dimension is lower than when we
included Moskos’ Institutional dimension. In examining our results across the four
models, we find that two dimensions of PSM (the higher an individual’s SS, the
greater the likelihood of being an ROTC cadet, and the greater their COM, the lower
their likelihood of being an ROTC cadet) and both 1-0 dimensions are related to the
likelihood of a respondent being an ROTC cadet. Perhaps, the findings indicate that
rather than PSM dimensions being a more nuanced measurement of Moskos’
Institutional dimension, they are picking up on a different set of motivations for
ROTC service.
D is c u s s io n a n d C o n c l u s i o n
As previously stated, budget cuts will likely affect salaries for military personnel
and may result in reductions in extrinsic motivators that might pose barriers to
452
Public Personnel M anagem ent 43(4)
recruitment, performance, and retention of personnel in the U.S. military. But as
Bright (2011) states, “public service motivation has the potential of transforming the
way employees are recruited, motivated and retained in public organizations” (p.
11). Our findings provide credence to the argument that the military may be able to
effectively diminish the role of monetary recruiting mechanisms. If the military con­
tinues to support policies in which they promote monetarily based recruitment tac­
tics, they may miss opportunities to recruit personnel motivated by intrinsic desires
to serve their country. This suggests that the military should consider who they want
to recruit as service people based on their motivations to serve the public.
We test three hypotheses in this article, two o f which are supported by our
analyses. We find that there is a positive correlation between Institutional motiva­
tions and CPV among ROTC cadets and that CPV is higher in ROTC cadets than
in the general undergraduate population. There is no support for our third hypoth­
esis, as CPV increases, so does the likelihood o f a respondent being an ROTC
cadet. Overall, three PSM dimensions (APS, CPV, and SS) are shown to be sig­
nificantly higher among ROTC cadets than among non-ROTC affiliated under­
graduate students (based on our difference o f means test). Overall, three PSM
dimensions (APS, CPV, and SS) are shown to be significantly higher among
ROTC cadets than among non-ROTC affiliated undergraduate students (based on
our difference o f means test); although as an individual’s COM score increases,
they are less likely to be an ROTC cadet. In addition, Moskos’ 1-0 dimensions are
statistically significant indicators of the likelihood o f a respondent being an ROTC
cadet.
This study is unique in its application of PSM to a population minimally studied in
the PSM field—those destined for military service. Although a significant body of
PSM literature exists among government and nonprofit employees both in the United
States and internationally, some gaps remain in testing the model across diverse popu­
lations of public servants. In addition, the findings indicate that combining the PSM
construct and Moskos’ 1-0 model may provide a more holistic picture of the underly­
ing motivations of military recruits.
Limitations
This study is limited in its generalizability because it is a convenience sample of
undergraduate students and ROTC cadets rather than a probability sample that is rep­
resentative of a university’s student body. In addition, the data are collected from a
Midwestern university and the responses collected may not be reflective of a broader
base of students from other colleges across the nation. That being said, given the
exploratory nature of this study, our interesting results suggest a need to replicate these
findings with more diverse samples to improve our confidence in generalizing the
implications of this research.
Ngaruiya et at.
4S3
Suggestions for Future Research
There are a number of themes that emerge for future research based on this study.
We unexpectedly f