 |
 |
 |
 |
|
 |
 |
|
 |
Internet use. The questionnaire was placed in the Opinion Place section of the
America Online member services screen. Individuals were randomly selected from
a pool of customer volunteers who were compensated for their participation with
free on-line time for use in their AOL accounts. The questionnaire was active for
this data collection over a one-week time period. This was a convenience sample,
even though subjects assigned to this task were randomly selected from a pool of
available respondents for projects that DMS was administering. Members of the
Opinion Place participant pool at AOL self-select themselves for participation in
researcher activities in exchange for a reward, so the construction of a sample from
this pool was not truly random.
A total of 343 usable on-line questionnaires were collected. The subjects
ranged in age from 18 to over 55, and included a mix of 165 men and
178 women. Fifty-
seven respondents were in the 18-24 age category, 72 were 25-34, 68 were 35-44, 70
were 45-54, and 76 were 55 or older.
ANALYSIS AND RESULTS
Initially, data on the 45 individual gratifications were subjected to principal
components analysis. Factor scores for the dimensions of Web use identified in the analysis
were then calculated and used as the independent variables in a multiple regression
analysis. In order to provide an opportunity to better understand which sort of motivations
were indicative of distinctive types of computer users (i.e., high use/low use), the affinity
for the computer, computer use frequency and Web use frequency measures were
specified as dependent variables.
Principal Components Analysis
The principal components analysis was conducted with varimax rotation, and
specified the retention of factors with eigenvalues greater than one. This resulted
in a total of 11 initial factors. Scree plot analysis (e.g., Hair, Anderson, Tatham, &
Black, 1998) indicated a natural break after the fifth factor, and of the five retained
factors, when reliability coefficients were calculated, it was noted that coefficient