METHODOLOGY                
   The focus of this preliminary study was primarily on Kirkpatrick’s third level of analysis and aimed at perceptions of actual changes in behavior that could be attributed to the training event.    A  detailed questionnaire was developed aimed at determining the effects of Internet training on life-style,  research productivity, earning power, management style, teaching and other factors.  Fifteen questions were used to  
  7 determine predictor or impact variables. 

These are shown in Table 1.
  It was expected that competency in World Wide Web use would be associated with perceptions of success in doing research and two questions probed this issue.  Second, three questions asked about the effect of Internet use on  organization  and management.  Since Internet skills can be used in the classroom, three questions were included  to determine  this  effect,  if  any.   We were particularly interested in the possible impact of Internet on the development and maintenance of teamwork structures in organizations and included four questions on that subject.  Finally, there were four questions that  asked about personal variables like the relationship between Internet and changes in job skills, earning power and quality of life.         
   The  questionnaire was sent by email to 521 persons who had completed the three-day course and 207 usable responses were received.  A sampling of the non-respondents indicated that there was no noticeable bias involved in the decision not to respond.  These non-respondents were relatively proportional to  the respondents according to age, academic discipline and gender.
 A response rate of nearly 40 percent for a questionnaire that requires over twenty minutes to complete is considered  excellent.   This  significantly exceeds the 20 percent response rate recommended for surveys of this type sent by mail [26]  and  the 6 percent rate for email studies reported by Tse [21].            

        Table 2 summarizes the population according to five individual variables: age, gender, research productivity, computer skills, and academic discipline.  Two thirds of the respondents are women, mirroring the demographics of the group that attended the Internet Academy courses