Mobile Edition
Print
Get the Mag
Weekly UpdatesANALYSIS AND RESULTS
As a first step, the item level data from the biodata instrument were subjected to a full-profile principal component factor analysis with a varimax rotation. The analysis produced five factors. Table 1 presents the five factors and the biodata items (including the Need for Achievement items) that had at least a .30 loading on the factor. Items were allowed to appear in more than one factor. Although it was the responses to the actual items which were entered into the factor analysis, what appears as the variables in Table 1 are statements that capture the sense of the actual items (the actual items are listed by McBride et al. (1997)). It would appear that Factors I through V may correspond to the Big Five factors of Extroversion (Factor I), Conscientiousness (Factor II), Emotional Stability (Factor III), Agreeableness (Factor IV), and Openness to Experience (Factor V).
The biodata items were then used to construct factor scores on each of the factors for each subject. In addition, subscale scores were calculated for each subject for each of the subscales identified by McBride et al. (1997). The correlations between each of the factor scores, each of the biodata subscale scores, the dependant variable (Service), Age and Gender are found in Table 2. As can be seen, the service orientation rating (Service) is significantly correlated with the factors of Extroversion, Conscientiousness and Openness to Experience. At the same time, the Service measure is significantly correlated with all the biodata subscales with the exception of Lack of Resistance to Stress. Moreover, with few exceptions, each of the biodata subscales correlates significantly with each of the factors. This last result should be viewed with some caution given that both the factor scores and the scale scores come from the same set of data. Therefore, for instance, the correlation between the factor Extraversion with the biodata scale of Sociability is .99 as the two come from nearly the same set of variables.
In the next step, the factor scores were regressed on the service orientation rating, and the biodata subscales scores were also regressed on the service orientation rating. Each of these analyses was done twice, once including age and gender and once excluding the demographic variables.
The results of the regression analyses with respect to the factor scores are presented in Table 3. As can be seen, the overall R is statistically significant and the [R.sup.2] is meaningful in both analyses. The largest individual contributor to service orientation rating is Extroversion. Openness to Experience and Agreeableness are also significant contributors. When included in the analysis, Gender has a small, but statistically significant relationship to the service orientation rating.
The results of the regression analyses with respect to the biodata scale scores are presented in Table 4. As can be seen, the overall R is statistically significant and the [R.sup.2] is meaningful in both analyses. The largest contributors to the service orientation rating are Good Impression, Agreeableness, and Sociability. These results are found regardless of whether age and gender is included. However, Gender does appear to have a small, but statistically significant relationship to service orientation.
DISCUSSION
This discussion will focus on three elements: the managerial implications of the results of the current study, the relation between these results and other studies using the Big Five personality factors, and suggestions for further research.
Managerial Implications
The results of this study suggest that a personality instrument, in the biodata format (McBride et al., 1997), might be a useful tool to consider for selecting individuals for service- related positions. As Table 2 indicates, all but one of the biodata subscales correlate significantly with the service orientation rating. This is reinforced by the results of the regression analysis (Table 4) where the biodata instrument, taken as a whole, accounts for over twenty percent of the variance in the service orientation ratings, and five of the seven subscales contribute to this overall effect.
When used in professional practice, the biodata instrument would have two advantages over the Service Orientation Index (Hogan et al., 1984). The most obvious is its length; the McBride et al. (1997) biodata instrument contains 39 items as compared with 87 items in the SOI. The second is that the biodata instrument has a stronger behavioral base than the SOI. This emphasis on a stronger behavioral base is recommended by Bowen et al. (1989).
It is worth noting that, despite the available literature, some writers are not supportive of the use of any type of service orientation test as part of the selection process. Fitzsimmons and Fitzsimmons (1998) recognized the importance of contact personnel in the service experience, yet argued that no reliable test existed to measure service orientation. They recommended a variety of interview and role-playing techniques as the preferred method of selecting service personnel.
Relation to Big Five Personality Factors
The results of the factor analysis (Table 1) indicate that the McBride et al., (1997) biodata instrument may be co-extensive with the Big Five personality factors. There was no a priori reason to believe that this result would be found, especially in light of the fact that the McBride et at. (1997) items did not evolve from a standard personality instrument. However, as the McBride et al. (1997) instrument is a personality measure, it should not be entirely surprising that it would decompose into the Big Five personality factors.
Once the factor scores were calculated, two interesting results were found. First, the service orientation rating was significantly correlated with Extroversion, Conscientiousness, and Openness to Experience (Table 2). With regard to Extroversion and Conscientiousness, this replicates the findings of both Barrick and Mount (1991) and Dunn et at. (1995) who found those two Big Five factors were related to sales performance or hireability judgements for insurance sales positions. The second interesting result is that a combination of the Big Five factors (at least as measured by the biodata instrument) has a high multiple correlation with service orientation and predicts a significant amount of the variance in that outcome measure.
This last result reinforces the recommendation that personality tests be considered for inclusion as part of the selection process for service-oriented positions (Bowen and Schneider, 1985; Bowen et al., 1989; George and Jones, 1991). However, we notice that the multiple correlations are nearly identical regardless of whether the factor scores or the biodata subscales are used. This would seem to suggest that the extra scales from the McBride et al. (1997) biodata instrument may not contribute unique predictive power above that achieved by just measuring the factors with the instrument. This is in contrast to Crant (1995) who found that, for real estate agents, a measure of Proactive Personality contributed to predicting performance beyond that achieved by Extroversion and Conscientiousness. We do believe that if this instrument is to be used as a measure of the Big Five that additional items be developed in order to measure Openness to Experience.
Suggestions for Further Research
A significant limitation of this study is that the outcome measure in this study was observed service orientation in a simulated customer service situation and not service behavior itself. Additional research using a predictive validity design and actual performance measures is recommended. This is especially important if an employer should have to defend the biodata instrument against an adverse impact claim under various employment discrimination laws (Black, 1994). This study was also limited by the fact that we did not have an independent measure of the Big Five personality factors and therefore it may be useful to collect information on the current scale and an existing measure of the Big Five and examine the convergent validity of the two scales. It could also be useful to examine how eliminating items from the scales might influence the scales' utility in order to minimize the number of items to effectively measure the Big Five.
It might also be advantageous to examine what other work place-related behaviors this biodata questionnaire may be useful at predicting. For instance it might be found that it could be useful for predicting turnover potential, attitudes towards absenteeism, or general work place performance (Buckley et al., 1998). An advantage of a biodata scale such as this one is that it is likely that the results will not be influenced by cognitive differences in job applicants which tend to be common with more traditional measures of personality (Carraher and Buckley, 1996). This should, however, be explored empirically.
This study examined the usefulness of a biodata instrument for the measurement of service orientation. The results indicate that the instrument is a valid predictor of judged service orientation in job applicants. They also suggest that the instrument may be closely related to the Big Five personality factors.
SUMMARY AND CONCLUSIONS
There has been increased interest in the use of personality instruments in the selection of personnel for positions with significant customer service responsibilities. One approach is to use a traditional inventory format. This study explores biodata as an alternative format for measuring service orientation. Applicants for a customer service position completed a biodata instrument and responded to simulated calls from customers. Service orientation was assessed by trained expert judges. The subscale scores of the biodata instrument were correlated with the service orientation assessment. All but one of the six subscales were significantly related to the service orientation assessment, and the multiple correlation between the subscales and the assessment was significant and accounted for more than 20 percent of the variance.
FEED