Abstract
The current study examined work value differentiation and profile elevation (PE). Using a sample of 251 college students who provided data on the O*NET Work Importance Profiler online, it was found that (1) when differentiation of work values was calculated using three indices (i.e., high–low differentiation, Iachan differentiation, and variance differentiation), only Iachan differentiation positively related to indecision; (2) none of the three indices of differentiation related to career maturity; (3) work value PE positively related to extraversion and openness and negatively related to depressive symptoms and career indecision but was unrelated to career certainty and neuroticism; and (4) work value PE moderated the relationship between Iachan differentiation and career indecision. The findings from this study have valuable implications for both vocational counselors and clients and should help to improve the utility of individuals’ work values results.
Keywords
Katzell (1964) defined a work value as “a stimulus or job characteristic which evokes a relatively high level of satisfaction” (p. 346). Research evidence suggests that work values are vital to consider, as the match between one’s values and those reinforcers in his or her work environment is a significant predictor of job satisfaction (Rounds, 1990). Given this, there are decades of research on work values assessment tools, one of which is O*NET’s Work Important Profile (WIP; McCloy et al., 1999). The WIP is a recent revision of the widely popular Minnesota Importance Questionnaire (MIQ; Rounds, Henley, Dawis, Lofquist, & Weiss, 1981). The WIP assesses 21 work needs that can potentially be reinforced in one’s work environment that are grouped into six overarching work values (i.e., achievement, independence, recognition, relationships, support, and working conditions).
Despite strong evidence of reliability and validity ( McCloy et al., 1999), the WIP report ignores information regarding the overall strength of scores or the profile elevation (PE) across the six work values assessed. This is problematic because, while two profiles may be considered similar in ranking of the specific values or needs, closer examination of results may reveal that they are very distinct profiles given differences in the absolute magnitude of scores. For example, two individuals may produce WIP profiles with the same ranking across the six values, yet one person may have endorsed all the values similarly high while a second person may have endorsed all values similarly low, while the within-subject hierarchy of values of the two individuals is the same. In other words, while the relative ordering of both individuals’ results is the same, the overall PEs are very different.
Two seemingly similar profiles may actually differ in another way given the dispersion of scores. For instance, two people may have the same values hierarchies (e.g., achievement, working conditions, recognition, relationships, support, and independence), but one person may have endorsed achievement and working conditions highly, recognition and relationships moderately, and support and independence minimally. In contrast, the other person may have the same order of values but have endorsed achievement slightly more than working conditions and endorsed working conditions slightly more than recognition, and so on, resulting in a narrow range of scores for the six values. In this case, their WIP profiles would appear similar in terms of ranking of values, but the dispersion across values scores is very different. In other words, the differentiation (D), or the variability of the scores, is different.
Although PE and D have not been examined in respect to work value profiles, application of these constructs to vocational interests suggests that they are relevant constructs. For instance, PE and D of vocational interests have proven to be related to individuals’ personality (Fuller, Holland, & Johnston, 1999), career exploration, career readiness, and vocational identity (Hirschi & Läge, 2007). Therefore, the PE and D of work values are worth examination as these subtle nuances of one’s work values results may provide helpful information about a client’s career development and improve the utility of the WIP for career professionals. Consequently, the current study sought to advance the literature on work values assessment by investigating how D and PE of individuals’ work values profiles related to personality and career development constructs. Previous research on D and PE applied to work values is almost nonexistent, leaving very few investigations to inform the present study. In contrast, these constructs have almost exclusively been examined with vocational interests. Given this limitation, review of research on D and PE applied to interests was done to provide a context for the likely relevance of these constructs to work values.
Holland’s Secondary Constructs: Differentiation and PE
Beyond discussion of six types of vocational interests (e.g., realistic, investigative, artistic, social, enterprising, and conventional [RIASEC]), Holland’s theory (1997) specifies a number of diagnostic signs (sometimes referred to as secondary constructs) to further understand one’s pattern of interests and correlated outcomes, including differentiation and PE. Differentiation (D) reflects the distinctiveness of a person’s profile of RIASEC themes, while PE refers to the sum of six RIASEC scores (Holland, Johnston, & Asama, 1994), capturing the overall magnitude of one’s interest results. Given that D and PE are constructs defined by Holland, they have been studied almost exclusively with interest data. Thus, in the following sections, the research on these constructs as they pertain to interest D and PE is reviewed to inform hypotheses regarding their applicability to work value assessment.
Differentiation (D)
Identified by Holland (1997), D describes how well a person’s interest profile is defined. A highly differentiated interest profile is typified by one dominant RIASEC interest type with relatively low endorsement of the remaining five RIASEC types. In contrast, low D describes a profile in which all the RIASEC interests are similarly endorsed. There are two ways to calculate the interest D noted in the literature. One is the high-low D, which is also called the traditional D score (Holland, 1997). A high–low D score can be obtained by subtracting the lowest RIASEC summary score from the highest summary score (Holland, 1997). The other method is Iachan D index, which is obtained by summing up the second and fourth highest RIASEC scores and dividing the sum by 2 and subtracting this resulting score from the highest summary score and multiplying this subtotal by .5 (Iachan, 1984). Furthermore, because D essentially describes the variability of scores in a person’s profile, researchers studying general values or work values, operationalize D of work values profiles using within-subject variance (Miller, 1974; Shrum & McCarty, 1992), which will be referred as the variance D throughout.
As an example of calculating the three different indices of D, take the following individual’s interest data where R = 51, I = 64, A = 57, S = 55, E = 46, and C = 42. The score for Iachan D, would be calculated as [64 – (57 + 51)/2] × .5 = 5. For high-low differentiation D, it would be 64 − 42 = 22, while variance D would be the variance of scores equaling 62.7. Some researchers have suggested that only the Iachan D index should be used for research on D of interests because it better captures the overall shape of the profile (Alvi, Khan, & Kirkwood, 1990), while variance D has been applied exclusively with work values. Overall, there appears to be some disagreement about the index that best reflects profile D.
Indecision
Holland (1997) originally proposed that low interest D may be related to higher career indecision because low D indicates similarly endorsed interests, and thus there are not one or two prominent areas of interest to guide one’s career choices. However, research has not consistently found such relationships (Holland & Holland, 1977; Lowe, 1981) or only found modest relationships (Jaensch, Hirschi, & Spurk, 2016). One explanation is that researchers have conceptualized and measured indecision inconsistently. For example, Holland and Holland (1977) conceptualized indecision as a state of being undecided on a career, measured via self-report items such as “I have made a tentative occupational choice or I am currently employed full time.” Similarly, Jaensch, Hirschi, and Spurk (2016) used My Vocational Identity Scale (Holland, Daiger, & Power, 1980) to measure indecision, which similarly focuses on undecidedness itself. In contrast, Lowe (1981) measured indecision using the Career Decision Scale (Osipow, Carney, & Barak, 1976), which assessed educational or vocational undecidedness via 16 antecedents leading to undecidedness, such as “I can’t make a career choice right now because I don’t know what my abilities are.” Thus, undecided may be operationalized as not having made a career choice, while career indecision refers to a dispositional tendency to have trouble making decisions, one’s inability to decide given limited information, or lack of readiness to engage in career planning (Osipow, 1999). Confusion regarding the nuances between these constructs, and that indecision is multidimensional, may be contributing to inconsistency about the relationship between indecision, undecidedness, and D in literature. Moreover, inconsistent findings may be due to the use of traditional high-low D, a weak index of D, as Alvi, Khan, and Kirkwood (1990) argued that Iachan D captures the overall profile better than the traditional high-low D and recommended using Iachan D in research instead.
Although past research indicates that interest D was unrelated to career indecision, work values D may still be related to career indecision because low D in work values reflects a state where the importance of various aspects of a job are less distinguished, which may consequently make it difficult to make a career choice. This supposition is consistent with Brown’s (2002) theory of occupational choice, satisfaction, and success, as he reasoned that the most crucial determinant of career choice is pursing one’s highly prioritized work values. In other words, when a person’s work values are prioritized, they also are highly differentiated. In contrast, when values are not prioritized, they are undifferentiated because the person is not clear about the relative importance of these work values and instead rate them as equally important. Therefore, it would be expected that D of work values negatively relates to career indecision.
Career maturity
Career maturity reflects how prepared a person is to make well-informed and developmentally appropriate career decisions when dealing with career development tasks (Savickas, 1984). Miller (1974) found that work values D and career maturity were moderately correlated for women (r = .43, 95% confidence interval CI [0.13, 0.66]) but not for men (r = .13, 95% CI [−0.29, 0.51]). These results suggest that low D, where values are endorsed as equally important, may delay one’s ability to make career decisions. Issues with this study’s sample, such as small sample size (24 men and 38 women) and the men in the sample being significantly younger and thus possibly less vocationally mature than the women, may have affected findings. These limitations question the generalizability of these findings and warrants further investigations on the relation between work values D and career maturity.
PE
Holland, Johnston, and Asama (1994) conceptualized interest PE as the overall level of one’s interests, calculated by summing the RIASEC scores. Alternatively, instead of using total scores, some researchers have used the one highest RIASEC score in the profile to operationalize PE (Swanson & Hansen, 1986). However, the sum of scores better captures the overall elevation of one’s profile than just examining the magnitude of the highest score, because two persons can have the same highest score, but very different sum of scores.
Personality traits
Fuller, Holland, and Johnston (1999) argued that people high in extraversion tend to be “outgoing, sociable, and cheerful” and people high in openness are inclined to be open to new experiences, including being open to new vocations and skills and therefore have more preferences and consequently produce more elevated interests. Supporting this assertion, interest PE has been found to be positively related to personality traits such as openness and extraversion and negatively related to neuroticism, although neuroticism was only significantly related to interest PE in men in some samples (Fuller, Holland, & Johnston, 1999; Gottfredson & Jones, 1993; Holland et al., 1994). Similarly, individuals with high openness and extraversion and low neuroticism may also show the tendency to be open about different kinds of needs and values and thus positively endorse the items they have not explicitly considered before in work values inventories and thus produce higher work value PE.
Depression
Supporting Holland’s (1997) supposition of low interest PE as indicative of depressive traits, interest PE has been shown to be negatively related to depression (Holland et al., 1994). This negative relationship is likely due to the reduction of interests and drives, which is one of the hallmarks of depression (American Psychiatric Association, 2013). Costello (1972) proposed that depression might result from loss of effectiveness of reinforcers. For people with depression, some aspects of the job that used to be reinforcers (e.g., achievement) may become less effective or even ineffective, leading to lower endorsement of importance of these work values. Therefore, it could be expected that work value PE also would be negatively related to depressive symptoms.
Career certainty
Career certainty is one aspect of the career decision-making process and describes how certain a person feels about one’s career choice. Schulenberg, Vondracek, and Kim (1993) found that groups endorsing moderate and high career certainty reported greater importance in work values related to self-development, power over self and others, and money and security than individuals in the low career certainty group. However, no differences were found in the importance of values pertaining to a preference for jobs in which disengagement from work is likely. Although this study did not explicitly examine work value PE, finding that the magnitude of values was higher for differing levels of career certainty provides initial support for the positive association between PE of work values and career certainty.
Interactions Between D and PE on Indecision and Maturity
Holland (1997) assumed that individuals with low interest D, as labeled undifferentiated, were a unitary group of individuals whose interests are not well defined and thus are likely to be unpredictable and have loosely defined goals and, consequently, low vocational and academic achievement. Yet research has not supported these assumptions. Instead, interactions between interest D and PE appear to differentially relate to career development, where PE moderates relationships between D and career development variables. For instance, Swanson and Hansen (1986) found that, compared to individuals with low interest D and low PE, individuals with low D but high PE had higher grade point averages (GPAs), academic comfort, and probability of staying in college. More recently, Hirschi and Läge (2007) found that middle school students with undifferentiated interests, but higher PE, reported increased career exploration, readiness, vocational identity, and career planning than those with both low D and low PE of interests.
One reason why students with low interest D, yet high PE, showed better outcomes may be that they had stronger interests, albeit undifferentiated and thus had a greater likelihood to find some aspects of their coursework enjoyable. Another explanation may be that some undifferentiated individuals could be multipotential, having skills and abilities in a number of different areas (Reardon & Lenz, 1998). These multipotential individuals tend to have abundant interests and talents, thus likely to have higher interests in multiple areas, typically generating a highly elevated, but undifferentiated profile (Rysiew, Shore, & Leeb, 1999).
Similarly, when studying the relationship of work values D with career indecision or maturity, it would be important to consider possible interactions between D and PE. To individuals with undifferentiated but highly elevated values, many aspects of the work environment are important and thus motivating. In contrast, those with low D and PE of work values may have lower motivation because they do not have as strong needs to fulfill through work. Moreover, high work value PE makes it more likely that at least some aspects of their important needs are fulfilled in a job. Therefore, high PE may reflect being more adaptive, and thus more mature regarding one’s career development process, which in turn may lead to less indecision. Despite no research examining if work values PE moderates relations between values D and career development constructs, there is theoretical support to hypothesize that among those with low D of work values, having higher PE may relate to lower indecision and higher career maturity.
Study Goals
Given the frequent use of work values assessments in the provision of career counseling services (Watkins, Campbell, & Nieberding, 1993), the current study aimed to advance the literature on work values assessment and interpretation. Specifically, we sought to determine whether work values D was related to career maturity and career indecision. In addition, we compared three indices of D—traditional or high-low D, Iachan D, and variance D—to examine which accounted for most variability in career maturity and indecision. We applied the research on interest PE to work values by determining whether PE of work values was related to personality traits, depressive symptoms, career certainty, and career indecision. Lastly, we examined interactions between PE and D of work values and their relationship to career maturity and career indecision. Our explicit hypotheses are listed below.
Method
Participants
Participants were 251 college students (157 women, 93 men, and 1 participant not self-identifying as either a woman or a man), from a public university in the southeastern United States, with a mean age of 20.9 years (SD = 3.92). Individuals were recruited using an online psychological research recruitment system (https://usm.sona-systems.com). Of the sample, 62.9% self-identified as White or Caucasian, 25.1% as Black or African American, 4.4% as multicultural, 4% as Latino, 2.4% as Asian American, and 1.2% as American Indian. Regarding class standing, 39% reported being first-year students, 22.7% being sophomores, 17.5% being juniors, 17.5% being seniors, and 3.2% being in college for 5 or more years. Roughly, half (51%) of the participants reported that they were currently employed, with 9.6% working full time (i.e., working 32 hr or more per week).
Measures
WIP
Developed by the U.S. Department of Labor, the WIP (McCloy et al., 1999) is the most recent version of the rank-ordered MIQ. A 7-point Likert-type scale (1 = least important and 7 = extremely important), instead of the original rank-order form WIP, was used because the rank-order form produces ipsative (i.e., individual rank-order data), which limits the ability to use the resulting data for some statistical analyses as well as not providing information on the absolute magnitude of one’s values. A 7-point scale specifically was chosen given evidence that a 7-point scale tends to produce better reliability, validity, and discriminating power compared to scales with fewer points (Preston & Colman, 2000). The measure is comprised of 21 questions each assessing importance of 21 work needs organized into six higher order values scales (achievement—2 items, working conditions—6 items, recognition—4 items, relationships—3 items, support—3 items, and independence—3 items; McCloy et al., 1999). The six work values scores were calculated by averaging scores of the work needs items under each corresponding value (McCloy et al., 1999). For example, the work value achievement has two corresponding work needs, ability utilization, and achievement. Thus, their average is used to obtain one’s score for the work value achievement. The final WIP values scores range from 1 (least important) to 7 (extremely important). Evidence of construct validity has been found through exploratory and confirmatory factor analysis, showing WIP had very similar factor structure as MIQ (McCloy et al., 1999). In an unpublished study (M. Leuty, personal communication, February 4, 2016), the Likert version WIP yielded reliability of .80 in a sample of working adults (N = 258). The reliability of WIP for the present study was .91.
Calculating PE and differentiation scores
PE was calculated by taking the sum of the six work values scores. The high-low D was calculated by subtracting the lowest value score from the highest. The variance D was obtained by calculating the within-subject variance for the six values using SPSS software (version 22). The Iachan D was calculated by adding the second and fourth highest value scores and dividing the sum by 2 and subtracting this score from the highest value score and multiplying this number by .5 (Iachan, 1984).
International Personality Item Pool (IPIP)
Personality traits were measured using 50-item from the IPIP (Goldberg, 1999), with 10 items for each factor of the five-factor model (i.e., extraversion, agreeableness, conscientiousness, neuroticism, and openness; Goldberg, 1999). The 50 items are public domain alternatives to the NEO Personality Inventory (NEO-PI-R: Costa & McCrae, 1992) and can be obtained online (http://ipip.ori.org/newNEODomainsKey.htm). Participants were asked to respond to items with a 5-point Likert-type scale (1 = very inaccurate; 5 = very accurate as a description of you). Factor scores were calculated by summing the scores of the 10 items for each factor resulting in a possible range of 10–50 for each factor with higher scores reflecting a higher level of the corresponding trait. The average reliability of IPIP is .80 (Goldberg, 1999). In the present study, internal consistencies were .89, .81, 85, .88, and .83 for extraversion, openness, agreeableness, conscientiousness, and neuroticism, respectively. Evidence for concurrent validity is available given high correlations, ranging from .85 for agreeableness to .92 for neuroticism and conscientiousness, between the IPIP and the NEO-PI-R (Goldberg, 1999).
Depression, Anxiety, and Stress Scales, Short Form (DASS-21)
The 7-item Depression subscale of the DASS-21 (Lovibond & Lovibond, 1995) was used to measure depression. Participants were asked to respond to items using a 4-point Likert-type scale, from 0 (did not apply to me) to 3 (applied to me very much or most of the time). Responses are summed to produce scores ranging from 0 to 21, with higher scores indicating greater endorsement of depressive symptoms. Osman et al. (2012) reported that among a nonclinical college sample the internal consistency for the depression scale was .85. The Cronbach’s α was .92 for the present study. Evidence for concurrent validity was established through the strong correlations between the DASS Depression subscale and other measures of depression (Antony, Bieling, Cox, Enns, & Swinson, 1998).
Career Decision Difficulties Questionnaire (CDDQ)
To measure indecision, the 34-item version of CDDQ (Gati, 2011) was used. The CDDQ measures the sources of career indecision or the domains of difficulties (Osipow, 1999). Items are organized into three main areas contributing to indecision, including lack of readiness (comprised of three subscales of lack of motivation [3 items], general indecisiveness [3 items], and dysfunctional beliefs [4 items]), lack of information (about the decision-making process [3 items], the self [4 items], occupations [3 items], and ways to obtain information [2 items]), and inconsistent information (including unreliable information [3 items], internal conflicts [5 items], and external conflicts [2 items]; Gati, 2011). As advised by the author, two validity items were disregarded when using CDDQ for research (I. Gati, personal communication, May, 2017). Participants were asked to respond to 9-point Likert-type scale (1 = does not describe me; 9 = describes me well). The average score across all items was calculated to indicate a person’s overall career indecision. Thus, the range of scores was from 1 to 9, with higher scores reflecting higher indecision. According to the manual (Gati, 2011), internal consistency of CDDQ ranges from .88 to .95 with a median of .93 across samples. Internal consistency of the total scale in the current study was .95. Evidence for convergent validity was provided by Osipow and Gati (1998), finding that CDDQ scores positively correlated with a measure career indecision and negatively correlated with scores on a measure of career decision-making self-efficacy.
Career Maturity Inventory (CMI) Form C
Career maturity was measured with CMI Form C (Savickas & Porfeli, 2011). The CMI is comprised of four subscales, Concern, Curiosity, Confidence, and Consultation, assessed with 6 items per subscale, for a total of 24 items. The participants responded either 1 = agree or 0 = disagree to each item. As suggested by the authors, the total score of three subscales, concern, curiosity, and confidence, was obtained to reflect a person’s overall career maturity because Savickas and Profeli (2011) recommend not to include the Consultation subscale in the total score due to its low correlation with the higher factor of maturity. Therefore, the possible scores range from 0 to 18 with higher scores indicating more maturity. Evidence for construct validity was provided by the adequate loadings of the three first-order factors (Savickas & Porfeli, 2011). In addition, evidence for concurrent validity was provided by the high correlation (r = .75) with CMI Form A-1, the version used with adolescents. Internal consistency of .86 for the CMI total score has been found (Savickas & Porfeli, 2011) and was .76 in the current sample.
Career Certainty
Based on brief literature search, only one standardized measure for career certainty was identified, which was comprised of two career certainty items of the Career Decision Scale (Osipow, Carney, Winer, Yanico, & Koschier, 1987). However, review of these 2 items found they were problematic as they were triple barreled, including statements that likely assess information outside of the construct of career certainty. Given the limited availability of psychometrically sound measures of this construct, it is common for researchers to use nonstandardized measures to assess career certainty (e.g., Daniels, Clifton, Perry, Mandzuk, & Hall, 2006). Thus, for the current study, career certainty was measured with the following six statements adapted from measures used in other studies (e.g., Daniels et al., 2006). “I have made up my mind on a career,” “I feel certain about my career choice,” “I am committed to my career decision,” “I have made up my mind on a major,” “I feel certain about the major I selected,” and “I am committed to my major.” Participants responded to these statements on a 7-point Likert-type scale (1 = does not describe me well to 7 = describes me well). An exploratory factor analysis suggested a two-factor solution which reflected career certainty and major certainty. The confirmatory factor analysis suggested an overall adequate fit for the two-factor model providing an initial support for the measure (root mean square error of approximation = .095, comparative fit index =.99, Tucker–Lewis index =.99). The internal consistency (.95) was high despite there were only 6 items in the measure. The two factors were also strongly correlated at .71. Based on these results, the 6 items were combined into a single score to assess the overall career certainty level, with higher scores indicating higher career certainty. Evidence for convergent validity was adequate given the moderate correlations with career indecision (r = −.57) and career maturity (r = .66).
Attention to Covariates
Review of the literature relevant to values assessment and the use of rating scales suggests private self-consciousness (PSC) and social desirability should be accounted for in the current study (Fenigstein, Scheier & Buss, 1975; Fisher & Katz, 2000).
PSC
PSC refers to the tendency to pay attention to one’s inner world, such as one’s thoughts and feelings (Fenigstein et al., 1975). Shrum and McCarty (1992) found that PSC scores were positively related to D of life values and concluded that individuals with higher PSC had higher D between different life values because they had better self-knowledge, awareness, and self-schema than individuals with lower PSC. Similarly, it is likely that this increased awareness as a result of higher PSC might not only relate to increased D in diverse life values but also relate to increased D of work values.
PSC was measured using the 9-item PSC subscale from the Self-Consciousness Scale Revised (Scheier & Carver, 1985) that includes a 4-point Likert-type scale (0 = not like me at all to 3 = a lot like me) applied to items such as “I’m constantly thinking about my reason for doing things.” High internal consistency (.75) has been found, as well as high test–retest reliability over a 4-week period (r = .76; Scheier & Carver, 1985). In the current study, the internal consistency was .77. The evidence for construct validity for the revised version was provided by Scheier and Carver (1985) who reported that the factor loadings and norms were largely the same as the original version.
Social desirability bias (SDB)
Crowne and Marlowe (1964) described social desirability as “the need for social approval and acceptance and the belief that it can be attained by means of culturally acceptable and appropriate behaviors” (p. 109), suggesting that individuals tend to present themselves in a favorable way, despite their true feelings about a subject or topic. This tendency is problematic because it may potentially suppress relationships existing between two or more variables (Ganster, Hennessey, & Luthans, 1983) or lead to spurious relationships. Fisher and Katz (2000) found that value measures contain a large portion of SDB in general and that the more important a certain value is in a culture, the more substantial SDB it will contain. Thus, scores from the 13-item Marlowe–Crowne Social Desirability Scale Form C (MC-C; Reynolds, 1982) were included as a covariate. In addition, by using confirmatory factor analysis, Loo and Thorpe (2000) found that MC-C showed adequate fit to the data, and supportive evidence of convergent validity with the original MC. Internal consistency of MC-C ranges from .62 to .76 (Loo & Thorpe, 2000; Reynolds, 1982) and 6-week test–retest reliability was .74 (Zook & Sipps, 1985). The reliability was .68 for the current study.
Procedure
Individuals who enrolled through the online recruitment system were directed to an online survey host (i.e., Qualtrics). After individuals consented to participate, the online survey questions were presented. Meade and Craig (2012) recommend having up to three validity items (i.e., “Answer this question as very true”) to screen out the participants who are not paying attention and responding carelessly. Therefore, three directed response items were intermixed with other questions in the survey. The survey took about 45–60 min to finish, and the procedure was approved by the university’s institutional review board. Individuals received partial class credit or extra credit in a psychology course in exchange for their participation.
Results
Of the initial 296 participants attempting the study, 27 men and 18 women were excluded because they failed one or more of the validity check items, had unrealistically short completion time (less than 10 min), or had more than 25% missing data. Therefore, a final sample size of 251 was used for analyses. Linear trend at point was used to impute missing data in two cases for the item “I am not interested in abstract ideas” from the IPIP (0.8% missing). Next, the means, standard deviations, and zero-order correlations between all study variables were calculated (see Table 1). Then, the assumptions for multiple regression were examined. Because the variance D and high-low D were highly correlated (r = .93), multicollinearity was an issue, although other assumptions of regression were met. As a result, separate regression models for each index of work value D were completed.
Alphas and Correlations Between Study Variables.
Note. αs for each measure are included on the diagonal where applicable.
a Not applicable.
** Correlation is significant at the .01 level (two-tailed).
Work Values Differentiation
For Hypothesis 1, multiple hierarchical linear regression was used to examine the prediction of career maturity and career indecision from the three indices of values D; high-low D, Iachan D, and variance D. Covariates were entered on Step 1, followed by one of the three D indices entered into the regression equation on Step 2. Results for career maturity, after accounting for PSC and MC-C scores, found that high-low D, ΔF(3, 247) = .05, p > .05, Δr 2 = .00; Iachan D, ΔF(3, 247) = 2.06, p > .05, Δr 2 = .01; and variance D, ΔF(3, 247) = 1.07, p > .05, Δr 2 = .00 were not significant predictors of CMI scores.
Results for career indecision, after accounting for PSC and MC-C, found that neither high-low D, ΔF(3, 247) = .00, p > .05, Δr 2 = .00 nor variance D, ΔF(3, 247) = 1.57, p > .05, Δr 2 = .01, were significant predictors of CDDQ scores. However, Iachan D, ΔF (3, 247) = 4.82, p < .05, Δr 2 = .02, β = .13, was a significant positive predictor of career indecision (i.e., CDDQ scores), although it was hypothesized that Iachan D would be a negative predictor. Thus, Hypothesis 1 was only partially supported.
Work Values PE
For Hypothesis 2, multiple hierarchical linear regression was conducted again to examine whether work value PE positively predicted extraversion, openness, and career certainty and negatively predicted neuroticism, depression, and career indecision. After controlling for covariates, the results showed that PE was a significant positive predictor of extraversion, ΔF(3, 247) = 7.04, p < .01, Δr 2 = .03, β = .17, and openness, ΔF(3, 247) = 9.51, p < .01, Δr 2 = .03, β = .19, as hypothesized but was not a significant predictor of career certainty, ΔF(3, 247) = 3.35, p > .05, Δr 2 = .01. Also, it was found that PE was a significant negative predictor of depression, ΔF (3, 247) = 11.03, p < .01, Δr 2 = .04, β = −.20, and career indecision, ΔF(3, 247) = 7.26, p < .01, Δr 2 = .03, β = −.17, as hypothesized but not a significant predictor of neuroticism, ΔF(3, 247) = .02, p > .05, Δr 2 = .00. Thus, Hypothesis 2 was partially supported.
Interactions Between Differentiation and PE
Because none of the three indices of work value D were significant predictors of career maturity, there was no justification to further test the Hypothesis 3 on the moderating effect of PE on the relationship between work value D and career maturity. Therefore, Hypothesis 3 was not confirmed.
For Hypothesis 4, multiple hierarchical linear regression was performed to examine the moderating effect of PE of values on the relationship between Iachan D and career indecision, given that Iachan D was the only significant predictor of career indecision. The method of examining moderation outlined by Frazier, Tix, and Barron (2004) was used. Before conducting the analyses, Iachan D and PE were centered. After entering covariates, Iachan D and the PE were entered into the regression equation on Step 2, although neither of the mains effects were significant (p > .05). On Step 3, the interaction term of Iachan D × PE was added. After accounting for PSC and MC-C, a significant interaction between Iachan D and PE, ΔF(5, 245) = 4.99, p < .05, Δr 2 = .02, β = .15, was found, as predicted. Results of this analysis are shown in Table 2. These results suggest that PE moderated the relationship between Iachan D and career indecision. Further probing this interaction, the simple slopes for 1 SD above and below the mean for PE were examined and plotted. CDDQ scores were more strongly related to Iachan D scores among participants with above average PE (+1 SD simple slope (SS) = 1.60, p < .001, 95% CI [0.72, 2.49]). In contrast, CDDQ scores were not significantly related to Iachan D scores among participants with low PE (−1 SD SS = .02, p > .05, 95% CI [−0.91, 0.95]) or average PE (mean, SS = .81, p > .05, 95% CI [−0.08, 1.7]).
Hierarchical Multiple Regression Results for the Moderating Effect of Work Value Profile Elevation on the Relationship Between Work Value Differentiation (Iachan) and Indecision.
Note. PSC = private self-consciousness; PE = profile elevation; Iachan D = Iachan index of differentiation.
*p < .05. **p < .01. ***p < .001.
Discussion
The findings suggest that differentiation (high-low, Iachan, and variance D) of work values is not related to career maturity but Iachan D appears to be a modest positive predictor of career indecision. In addition, work values PE was a modest positive predictor of extraversion and openness and a negative predictor of depression and career indecision while being unrelated to career certainty and neuroticism. Moreover, a significant interaction was found between Iachan D and work value PE in their relationships to career indecision, such that Iachan D is positively related to increased career indecision when PE is above average.
Work Values Differentiation
Finding that high-low D and variance D were not significant predictors of career indecision is consistent with the previous research on interest differentiation that has constantly failed to find evidence for relationships between the two. Moreover, Alvi et al. (1990) recommended using Iachan D index for research on D because it captures the shape of the profile better than traditional high-low D index. Their recommendation was supported by the current study given that Iachan D index did show a significant relationship with career indecision while high-low D and variance D indices did not, suggesting the index chosen to examine D in work values does make a difference. In contrast, none of the three indices of D were significant predictors of career maturity, suggesting that they are unlikely related.
Interestingly, differentiation was originally hypothesized to have a negative relationship with indecision because more refined work values are thought to facilitate career decisions (Brown, 2002). However, a positive relationship was found. One potential explanation is that with more differentiated values, people experience more difficulties in deciding on a career, maybe because matching values are limited. In other words, because they are looking to fit a specific hierarchy of values, it becomes more difficult to identify a specific occupation that reinforces these values. Moreover, those with less defined values, who express that all values are equally important, endorsed less indecision possibly because of the increased number of career options that may reinforce at least one or some of those needs. Put another way, it may be that highly differentiated values may actually decrease flexibility in selecting a career path. This is partially supported by the trending negative correlation between differentiation and openness, for all three differentiation D indices and openness (Table 1).
Work Values PE
PE of work values, similar to interest PE, was a significant positive predictor of openness and extraversion and a negative predictor of depression. This result suggests that elevation of work values, similar to elevation of interests, signals individuals who are more open to experiences (openness), outgoing (extraversion), and report few depressive symptoms. In contrast, PE of work values was not a significant predictor of neuroticism and career certainty. For neuroticism, the result is consistent with previous literature on interest elevation and neuroticism finding that the two are either modestly correlated (Gottfredson & Jones, 1993) or not significantly correlated at all (Fuller et al., 1999; Holland et al., 1994). Thus, it is concluded that work value PE, similar to interest PE, is unrelated to neuroticism.
As for career certainty, the previous research has found associations between work values PE and career certainty (Schulenberg, Vondracek, & Kim, 1993), yet this was not found in the current study. The inconsistent findings may be due to the lack of consistency in the measurement of career certainty and researchers’ use of their own measures, which make the comparisons between studies difficult. Favorable psychometric evidence for the use of 6 items used to assess career certainty was found. More studies using these items in different populations and contexts are required to further clarify the relationship between career certainty and work value PE.
As expected, work value PE was negatively related to career indecision, suggesting that if a person regards many work values as highly important, less indecision is also reported. One explanation is that when work value PE is high, individuals may feel that some of the most important work needs are likely to be met, which reduces difficulties in making career decisions.
Interactions Between Differentiation and PE
Work value PE moderated the positive relationship between Iachan D and indecision. More specifically, D appears to be related to indecision only when work value PE was at or above average, but they are unrelated when work value PE was at or below average. This result fits well with the speculation that one possible reason why Iachan D and indecision were positivity related is that highly refined work values may reduce the flexibility in identifying possible career paths, leading to career indecision. When highly defined values are also highly endorsed, these individuals may be more inflexible or more passionate about these particular values being met by their future career, which may lead to more indecision. Alternately, when values are not differentiated, yet highly elevated, less career indecision is reported. This result is consistent with previous studies on interactions between D and PE of vocational interests (Hirschi & Lage, 2007; Swanson & Hansen, 1986), reporting that among subjects with low interest differentiation (undifferentiated subjects), those with high PE, compared to those with low PE, reported higher career readiness identity, decidedness, planning, and career exploration. The results imply that higher elevation is actually helpful for both values and interests even when they are not differentiated because endorsement of many interests and values increases the likelihood of finding a career path that fits some of these preferences.
Lastly, both work value D and PE were significantly related to indecision in earlier analyses, but when D and PE of work values were examined together, main effects for both were no longer significant. Thus, the results indicate when considered together, it is the interaction between D and PE that drives the relationship with indecision, not D or PE independently, and suggest that consideration of both elevation and differentiation of work values is needed to understand how they affect career decision-making.
Implications for Practice
Findings suggest that indecision is closely related to both work value differentiation (Iachan index) and PE. Firstly, the positive relationship between D and indecision does not necessarily mean that having refined values is disadvantageous. In contrast, it can be advantageous because prioritized values help to narrow down possible career options to consider and may eventually lead to better matching careers and consequently higher job satisfaction. However, this advantage is based on the premise that the person is aware of his or her most important set of needs and utilize this knowledge to find matching careers by utilizing resources such as O*NET (www.onetonline.org; National Center for O*NET Development, 2018) to generate a list of jobs matching his or her specific values profile. Moreover, when individuals report highly defined values, which are also highly endorsed, these individuals may need career guidance. As such, for those with well-defined work values, career counseling may help to develop awareness about one’s values and connect one’s most important needs to possible careers (Brown, 2002). Secondly, when people report low differentiation among values, having more elevated values can guard against potential indecision. In contrast, individuals with low D and also low PE are likely to need the most assistance. These clients may benefit from additional mental health screening to rule out possible depressive symptoms and then counselors can help clients realize some important aspects of the work by exploring the reasons behind low endorsements of work values statements, such as lack of knowledge about oneself or the world of work.
Implications for Research
The current study is the first study that comprehensively examined work value differentiation and PE of work values. Therefore, more research is needed to verify the findings. Future studies also should investigate whether D and PE of work values moderate the relationship between work values congruence (i.e., the match between person’s values and the actual jobs) and job satisfaction. Although increased D may contribute to increased difficulties in the initial career planning process, it is still unknown whether more refined work values will eventually lead to better match between person’s values and the actual jobs and consequently higher job satisfaction. Another possibility is that people with highly differentiated values may need to have the “perfect” fit in their job to be satisfied and thus will have less congruence and consequently lower job satisfaction. Also, it seems reasonable to infer that work value PE will moderate the relationship between work values congruence and satisfaction because people with more highly elevated values are more likely to find jobs that match some of their values and thus are more satisfied at work.
The finding that only Iachan D had a significant relationship with indecision while high-low D and variance D did not, suggest that the Iachan D is probably a better index to assess differentiation, than traditional high-low D, probably because high-low D does not account for the scores in the middle between the highest and lowest value scores while Iachan D utilizes the highest, second highest and fourth highest scores and thus better captures the shape of the overall profile. The variance D is, in theory, the perfect reflection of D and it is unclear why a relationship with indecision was not found as predicted. At this point, the current results suggest that Iachan D is more related to indecision than other two indices of D. Thus, in the future, research on values and even interest differentiation should consider using the Iachan D index.
Lastly, the utility of the WIP Likert form in research is evidenced by meaningful relationships found in the current study. However, correlations between the Likert form and the traditional rank-order form should be examined in the future as evidence for concurrent validity.
Limitations and Future Research
One limitation is that the data were from college students, which could have led to restriction of range of certain variables. For example, only about half of the participants reported being employed. Consequently, they may only have limited work experiences and thus have not yet solidified their work values through work experiences, resulting in unstable work values. Additional research with noncollege students is needed to further advance this area of study.
Moreover, this is a correlational study and as such, findings do not tease apart whether differentiation or elevation of work values causes indecision. Longitudinal studies with high school and college students, such as the study by Hirschi (2009) discovering factors (e.g., gender, personality traits, and career exploration) contributing to the development of the interest differentiation and elevation, may help clarify the causal relationships. In addition, longitudinal data collected during counseling practice will provide strong evidence showing how D and PE of work values as well as career related variables, such as career indecision and maturity, change during the course of counseling, and how these variables relate to each other in real-life settings.
Conclusions
This study was the first systematic examination of PE and differentiation of work values and found that D calculated by the Iachan index was positively related to career indecision. PE was positively related to extraversion and openness and negatively related to depression and career indecision. Furthermore, the interaction between D (Iachan D) and PE suggests that D was related to increased career indecision only when PE was also high. In sum, the current study garnered support for consideration of D and PE as meaningful constructs for work values that warrant further study.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.