Social Skills Training for Cooperative Education and Internship Students: An Empirical Investigation of Performance Outcomes

There is no accomplishment so easy to acquire as politeness and none more profitable. (George Bernard Shaw).

During the last decade, the con­cept of social skills has received significant attention in the business and educational communities. Generally defined as the ability to effectively and efficiently interact within an interpersonal environment, social skills have also been regarded as an increasingly important com­ponent of the career success equa­tion. This is due, in large part, to the presumed influence that social skills have on a variety of work­ related performance outcomes. As one social skills expert recently explained, "People are just appalled at the lack of manners in today's business world. Companies have learned that when their staff has good etiquette, people aren't as mad at them, and they don't lose business." (Baldridge, as quoted in Roberts, 2000, p. 14)

In order to better understand the influence of social skills in the career success equation, one need only refer to the results of recent social skills surveys. For example, the National Association of Colleges and Employers recently reported that 44 percent of the employers who responded to its survey said that it is a job candi­date's general demeanor that first gets their attention ( Christian Science Monitor, 2000). From a customer relations perspective, a recent Eticon survey found that 58 percent of the respondents said they would take their business elsewhere when confronted with rude behavior (reported in Humphries, 1999). Finally, Dobbs (1999) notes that the Research Institute of America reports that "96 percent of cus­tomers never complain about dis­courteous employees, yet 91 percent make a concerted effort not to do business with a company that has slighted them." (p. 48) These and other survey results suggest that social skills can have a significant influence on performance.

As such, it should not be surpris­ing to learn that educational and business institutions have acceler­ated efforts toward providing for­malized social skills instruction. For example, the University of Virginia, Pennsylvania State University, Purdue, and the University of Denver are now offering seminars in dining etiquette (Joyce, 1998). Some of the more prestigious uni­versities are even sending their students to vintners such as Kendall-Jackson where they learn the art of wine tasting and "wine s­peak." The rationale is that "The best place to do business is over a nice lunch or dinner and it's good to know that you don't order a sauterne with a steak." (CEO Jack Cakebread, as quoted in Sinton, 1999, p. D2)

The importance of social skills training has also caught the attention of the corporate world. Neiman- Marcus, for example. has added an eti­quette seminar to its in-house training regiment (Hopper, 1999). Others, such as IBM, Marriott, and Merrill Lynch are sending some of their young executives to "business golf' seminars where they learn the art of selling on the golf course. The reasoning is that "Rounds of golf are now the 'martini lunches' of a generation ago, so companies need salespeople with social graces and tactical savvy on the links" (Weinbach, 2000, p. BI). Such training is particularly important since, "golf has a lot of little no-no's that people might use to judge your professional character" (Rhame, as quoted in Weinbach, 2000, p. B 1 ).

The perceived importance of social skills training has also resulted in the emergence of numerous consulting firms. At Ease, Inc., for example, offers a group training seminar that is promoted as pro­viding business executives with the "polish that builds profits." And it isn't cheap. The cost for a one-day session is currently in the neighoorhood of $3,500 if you travel to their the home office in Cincinnati. Ohio and approximately $6,000 if they travel to you.

While this cursory review suggests that the business and educational communities believe that social skills training will result in positive performance outcomes, no formal study could be identified that has formally examined this issue within the context of the traditional work environ­ment. Instead, the most common emphasis of social skills research has tended to focus either on topics not directed related to training or on topics not directly related to the traditional work environment.

With regard to non-training related social skills research, prior studies have shown that interview­ers who possess higher levels of "social perception" do a better job of choosing the best job applicants (Eder, 1989). Other research has shown that indi­viduals who possess relatively high levels of "impression management" skills are more likely to receive job offers (Stevens & Kristoff, 1995), as well as more likely to receive higher annual performance reviews (Giacalone & Rosenfeld, 1989). It has also been shown that individuals high in "social adaptability" attain "greater success and more rapid promotions in many different con­texts than those low in such abilities." (Baron & Markman, 2000, p. 108; Kilduff & Day, 1994). Finally, "entrepreneurs whose companies are suc­cessful engage in more communications with others, and are more effective in this activity, than entrepreneurs whose companies fail." (Baron & Markman, 2000, p. 114; Duchsenseau & Gartner 1990).

With regard to training-related research, prior studies have examined the effects of social skills training in such diverse areas as the development of self-concepts of gifted high-school students (Clark & Dixon, 1997), the treatment of childhood social phobia (Spence & Donovan, 2000), the acquisition of social leisure skills among adults with moderate levels of mental retardation (O'Reilly, Macionis, & Kierans, 2000), and the social responsiveness of dual sensory impaired employees (Gaylord & Park , 1995).

Given the presumed importance of social skills in the work performance equation, the apparent lack of research specifically devoted to evaluating the effects of training on work- performance eval­uations represents a critical gap in the literature. Does social skills training influence work-related performance evaluations? This study seeks to answer this question by examining the effects of a social skills training workshop on the performance evaluations of students enrolled in cooperative education/internship programs.

The general hypothesis is straightforward: stu­dent interns who receive social skills training will receive higher work-performance evaluations for outcomes that are directly related to the training. Conversely, since social skills training is not directly related to the development of technical aspects of work assignments, it is hypothesized that there will be no statistically meaningful differences between the treatment group (i.e., student interns who received the training) and the control group (i.e., student interns who did not receive the training).

The following sections provide a more complete discussion of the method of analysis, sample characteristics, treatment, and test results.

Methods

A one and one-half hour social skills training session was incorporated into the Spring 2000 semester orientation for students enrolled in coop­erative education and internship programs at a large land grant university. (Since most of the students were in internship programs, the phrase "interns" will be used in this paper to refer to the partici­pants.) The training session took place during the first part of the semester. Although attendance was not mandatory, students were encouraged to attend via a formal letter to the student's residence (sent two to three weeks prior to the scheduled orienta­tion). Additional notification was provided through campus bulletins, posters, and flyers. To further encourage student participation, two training ses­sions were offered; one in the early afternoon and one in the evening. A review of attendance indicated that of the 200 student interns enrolled during this term, 117 attended the training session. Therefore, the 117 intern students who received the training were designated as the treatment group and the 83 intern students who did not receive the training were designated as the control group. So, while only some of the students received social skills training, they all participated in an internship.

Given the quasi-experimental nature of the design, an important step in the evaluation process was to examine the extent to which the non-random assignment of students might have affected sam­pling error. To address this issue, a student char­acteristic profile was developed for each group. A comparison of the frequency distributions across three key characteristics provided information regarding the degree of discordance between the two groups. It was believed that highly discordant characteristic profiles would be evidence of sam­pling bias that could potentially affect the test results. More specifically, should the profiles be shown to be highly discordant, that would mean that the data analysis would need to be evaluated with a higher degree of caution. Alternatively, should the profiles be shown to be relatively proportional, that would indicate that the analysis could be evaluated with a greater degree of confidence.

Evaluation of Characteristic Profiles

Table 1 presents characteristic profiles for each of the groups evaluated in the study. Although a more complete characteristics profile would be helpful, the profile cards submitted by the students contained only the major demographic information categories of gender, major, and class level. As the table indicates, the profiles appear relatively proportional.

Treatment: Social Skills Training

The treatment consisted of an approximately 1.5 hour training session on work-related social skills. During the first part of the workshop, stu­dents were given a short true/false test to help determine their knowledge of social skills and to sensitize them to the fact that social skills knowl­edge is not always intuitively obvious. The fol­lowing are representative examples:
True or False?
Statement: A name tag should always be place on the left shoulder.
Answer: False, it should be worn on the right since that is where people's eyes are directed when shaking hands.
Statement: Not answering an RSVP is the worst thing one can do with an invitation.
Answer: False, not answering and then showing up is the worst. Most hosts and hostesses assume that if you don't answer, you are not planning on attending.
Statement: At a business reception, you should always hold your drink in your right hand.
Answer: False, the beverage should be held in the left hand to leave your right hand free for handshakes
Statement: It is ok to salt your food before tasting it.
Answer: False, it sends the message that you pre judge the world around you.

The second part of the workshop consisted of a summary review of the status of social skills in business today. The point of this section was to sensitize students to the fact that technical skills are not enough in today's competitive work envi­ronment; they need adequate social skills as well. As one social skills expert recently explained:
"There used to be a time when technical skills could distinguish you from your competitor. Today, technical expertise is assumed, and what distin­guishes [you now l is how well you present your­self." (Thomas, as quoted in Scott, 1999, p. 18) This part of the workshop also reported on the findings of several surveys which have shown how a lack of social skills can influence business per­formance. This was intended to illustrate that social skills, although relatively intangible, can influence tangible outcomes.

The third part of the workshop consisted of presenting selected aspects of work-related social skills knowledge. In order to meet this objective in an efficient manner, I used the classification out­line recommended by Bartkus (2001). In general, social skills knowledge was defined by numerous general categories (e.g., general office communi­cations). In turn, each of the major categories was further defined by several sub-components that were used to provide procedural social skills knowledge. For example, with regard to negotiation skills, students were reminded that words have been shown to communicate only 7 percent of the con­tent of the message, the rate and inflection of speech communicates 38 percent, and the nonverbal cues (body gestures) exhibit 55 percent (as reported in Scott, 1999). In doing so, students were sensitized to the potential influence of non-verbal communi­cations. A brief follow-up discussion on selected nonverbal techniques was also presented. As with all workshops, students were encouraged to become involved in the discussion and provide feedback on the material.

The fourth part of the workshop consisted of a summary of the material and a reminder that each organization has its own unique set of social norms and that an adaptation of these norms is part of the natural socialization process.

Measures of Evaluation and Hypthesis

At the conclusion of the internship, each stu­dent's supervisor provided a written evaluation of the student's performance during the internship. The evaluations covered eleven dimensions of work performance. Six of these were directly or tangentially related to social skills training and were therefore hypothesized to be affected by the training. Four of these were not directly related to social skills and were therefore not hypothesized to be affected by the training session. The remaining evaluation was an overall evaluation of the student intern. Since this evaluation was a global measure that encompassed both social and technical aspects of work performance, it was not hypothesized to be affected by social skill training.

H₁: The following six socially-related performance evaluations will be higher for the treatment group rather than for the control group. All evaluations were measured on a scale of one to six:

1. Interpersonal Effectiveness in the Workplace: defined as the extent to which the student intern is perceived to cooperate effectively with others in the workplace.
2. Ability to Handle Public: defined as the extent to which the student intern is perceived to be able to deal effectively with the outside public.
3. Promptness: defined as the extent to which the student intern arrives for work on time.
4. Responsibility: defined as the extent to which the student intern is perceived to accept responsibility.
5. Application: defined as the extent to which the student intern is perceived to be enthusiastic about the work assignment.
6. Initiative: defined as the extent to which the student intern is perceived to need supervision.

H2:The mean for the following four performance evaluations and the overall evaluation will not differ between the treatment group and the control group. All evaluations were measured on a scale of one to six.

1. Knowledge of Work: defined as the extent to which the student intern is perceived to have knowledge about the work assignment.
2. Work Accuracy: defined as the extent to which the student intern is perceived to complete a work task without "mistakes."
3. Quantity of Work: defined as the amount of work that the student intern is perceived to be completing.
4. Possibilities for Promotion: defined as the extent to which the student intern is perceived to have potential for promotion within the sponsoring organization.
5. Overall Evaluation: defined as a global measure of performance evaluation, the overall evaluation is intended to be a measure of the overall qualities of the student intern. Since social skills training is only expected to affect part of this evaluation, there is no strong rationale to suggest that it will have a significant influence.

Test Procedures and Results

At-test procedure was used to evaluate the effect of the social skills training session for each of the eleven hypotheses. The results are presented in Table 2. For the items where the mean evaluation scores were hypothesized to be different, three were shown to be statistically significant (i.e, p<.05) and one was shown to be marginally significant (i.e., p<.10). Specifically, the mean evaluations scores were higher for: Interpersonal Effectiveness in the Workplace, Promptness, and Application. Since the mean difference for Responsibility was shown to be only marginally significant, it should be viewed as preliminary at this point. However, the fact that the mean difference was in the hypothesized direction is encouraging. The test results for mean differences in Ability to Handle Public and Initiative were not significant (i.e., p>.10).

With regard to the performance evaluation items where the hypothesis of no mean difference was proposed, all were supported. That is, there were no statistically meaningful differences between the control and treatment groups for Knowledge of Work, Work Accuracy, Quantity of Work, Possibilities for Promotion, and Overall Evaluation.

Discussion

This study is one of the first to report on the effectiveness of social skills training on perform­ance outcomes in a traditional work environment and the first to report on the effects in a coopera­tive education or internship context. In general, the results support the argument that social skills training increases performance evaluation scores for those outcomes that are directly related to the training. It appears, therefore, that social skills training can have beneficial effects for cooperative education internships and for sponsoring organi­zations.

Since social skills training is not ordinarily part of formalized instruction in higher education, the implementation of a training program within the parameters of a cooperative education or internship setting may be a helpful step in filling this gap.

However, since two of the six hypotheses for directly related outcomes were not supported (i.e., Ability to Handle Public and Initiative), additional discussion is needed to help define the outcome parameters of social skills training.

With regard to Ability to Handle Public, the lack of a significant difference is a bit disconcerting. One possible explanation is that this "ability" is more complex than previously thought and may require more training than can be provided during a short one and one-half hour session. Alternatively. it is possible that the diversity in work assignments was such that some interns had significantly greater interaction with the public than others. Since the work files for the interns did not explicitly indicate the extent of public: involvement, this explanation is tentative. However. the relatively high number of students enrolled in information systems internship programs suggests that this explanation is plausible.

With regard to Initiative, it might be that the concept is not as directly related to social skills training as previously thought. Given that the hypotheses for a related outcome (i.e., "responsi­bility") was only marginally supported suggests that social skills training is not effective in improving initiative and responsibility and/or that the training session did not sufficiently incorporate those skills that are specifically aimed at fostering initiative and responsibility. A review of the session content does suggest that the concepts of initiative and responsibility were inferred from the training but were not as heavily emphasized as other issues such as interpersonal effectiveness. Therefore, it could be argued that the lack of statistical significance was due. at least in part. to the nature and scope of the training Future research may want to increase the emphasis on social aspects related to initiative and responsibility in order to help deter­mine the extent to which such skills training would affect performance outcomes.

The fact that there were no mean differences for the evaluative items that were not considered directly related to social skills training provides some discriminant validity for the results of the quasi-experimental design. In essence, had differ­ences been found for these additional items, it would have been more difficult to argue that the results were due to social skills training. Instead, the more likely explanation would have been that the results were at least partially due to sampling bias emanating from the non-random assignments. In this study, the ability to distinguish between related and non-related evaluative measures pro­vided evidence of such validity.

As with all research, it is important to discuss some limitations of the current study and to provide directions for future research. First, future research should strive to provide a random split for the two groups. This was difficult in the current study since the process of informing students about the orien­tation session meant that a successful random split would have been difficult to obtain. It is anticipated that other work-integrated programs might face similar obstacles. Perhaps more importantly, there is a potential ethical issue of providing some stu­dents with skills training but not others. That is, since each student has the right to equal treatment, excluding some students solely on the basis of research may not be sufficient. This is, needless to say, one of the more problematic issues sur­rounding the application of an experimental design in the context of a real life environment.

It could be argued, however, that the need to validate the effectiveness of social skills training over-rides the perceived dilemma of excluding some students from participating in the training session. Specifically, because of the increased importance that organizations and academic institutions are placing on social skills, there is a need to validate not only whether or not the training is, effective. but what types of training affect what types of outcomes. Future research that validates the results of this study using a random split would be a useful first step towards resolving this issue.

With regard to the validity of the current study, it is important to note that it is not possible to say with a high degree of certainly that sampling bias did not affect the test results. All that can be said is that there is no noticeable evidence that the two groups differed with regard to their characteristic profiles. It is still possible, however, that students who voluntarily attended the training session had a greater need for such training. Of course, had this been the case, there would likely have been a down­ward bias on the mean differences. That is, those who attended one of the sessions may have been less socially developed and in greater need of social skills training. If this is the case, the test results should be considered conservative and robust.

Alternatively, it may be that the students who attended the training session were more responsive to social pressures and perhaps more sensitive to the social environment than those who did not attend the session. In this case, the impact on the test results would more likely be upwardly bias. In other words, if interns who attended the session were more responsive to the social environment in the first place, it could be argued that their overall social skills were more developed as well and this might explain the mean differences in the test results. A replication that utilizes random assign­ments would help resolve the issue.

Finally, one might ask whether cooperative education or internship students differ from other students with regard to their level of social devel­opment. Although not part of the current study. additional research in this area may provide impor­tant information. For example, given that cooper­ative education or internship programs are often elective, one might ask whether or not students who participate in these programs recognize a need for additional development. The central research question here is whether or not the student's "need ot participate" is the result of perceived deficiency in social development compared to other students or simply excel beyond the norm. Future research that helps answer this question would enhance our understanding of the relationship between social skills and performance evaluations.