ACTION's 1991-92 VISTA Marketing Campaign: An Analysis. STUDY DESIGN


The VISTA marketing campaign consisted of two major components: a recruitment effort and an advertising campaign. The VISTA marketing campaign came out of ACTION headquarters in Washington, DC. Other VISTA marketing activities were occurring at the same time as part of normal activities from the state ACTION agencies and local sponsors.

Forty-five (45) college campuses and universities were selected for the marketing effort--five campuses from each of the nine ACTION Regions. The recruitment effort began in September 1991, when VISTA recruiters started full-time activities. These activities included talking to students either individually or in groups, putting up posters, and distributing brochures. The advertising campaign, which began in January 1992, primarily focused on print media, such as magazines and campus newspapers.

A quasi-experimental design was developed to evaluate the effectiveness of the marketing strategy. In each ACTION Region, one Test (or experimental) campus was selected from the five marketing campuses. Each Test campus was matched to a similar Comparison campus selected from the non-marketing campuses. Matching was done on the basis of public-private status, enrollment size and geographic location. The idea is that unique features of a Test Campus are likely to be found on its counterpart Comparison Campus, so that differential effects of marketing are not likely to be due to these features.

Student Surveys were conducted on the nine pairs of Test and Comparison campuses. The Student Surveys were conducted by telephone at two points in time. The marketing plan called for the main thrust of the recruitment campaign on the Test campuses to be conducted in the Fall of 1991 and the Spring of 1992. The campaign began, after some delays, in January 1992. The first Student Survey was to be conducted in both the Test and Comparison campuses in January 1992 as a pre-campaign interview with the second Student Survey to be conducted during April and May 1992 on these same campuses as a post-campaign survey. The Student Surveys were timed to occur one week before final exams.

One-hundred (100) randomly selected Juniors and Seniors on each of the nine Test campuses and the nine Comparison campuses participated in the first or pre-campaign Student Survey. A different sample of 100 randomly selected Juniors and Seniors was chosen to participate in the second or post-campaign Student Survey. Thus, altogether there were approximately 1,800 precampaign Student Surveys (900 each in the Test and Comparison campuses) and 1,800 post-campaign Student Surveys (900 each In the Test and Comparison campuses).

As in any sample survey, there are potential sources of error, principally sampling error and measurement error. In this study, sampling errors may be attributed to differences in the surveyed and non-surveyed campuses and in the students selected from these campuses. While we are dealing with matched test-comparison campuses, unobserved differences rather than the marketing intervention could affect the results.

Measurement errors may arise from factors like: (a) unreliability of answers from respondents due to faulty memory, phrasing of questions or unobserved processes, and (b) differences in the manner of survey administration from campus to campus and time period to time period. Specifically, in some Test campuses, the marketing campaign began before the pre-campaign Student Survey was conducted. In particular, VISTA recruiters began their activities earlier than anticipated. Hence, any pre-post comparison may understate the effect of the campaign in those cases. Furthermore, since the VISTA marketing campaign was only part of overall VISTA recruitment activities, since there were differences among recruiters in their specific recruitment activities, and since some overall recruitment activity could have occurred on Comparison campuses, this was not a true experimental design in the classic sense.

Finally, data from one Region were excluded from pair-wise comparisons, because a Comparison campus in that Region was found later to have been the site of test marketing efforts. Data from the corresponding Test campus in that Region were also excluded.

Information was also used from the VISTA Applicant Survey, a mail survey of VISTA applicants who had either recently graduated from college or were upperclassmen. In all, 504 VISTA applicants responded to the survey. In addition, a VISTA Recruiter Survey was administered by telephone to nine recruiters on Test campuses.

View full report


"action.pdf" (pdf, 32.84Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®