ERIC Number: ED349868
Record Type: Non-Journal
Publication Date: 1992-May
Reference Count: N/A
Utilizing Throw-Away Data: Invalid and Missing Data Can Have Meaning! AIR 1992 Annual Forum Paper.
Smith, Constance K.; Waggener, Anna T.
Two studies were conducted at the Southeastern Louisiana University (SLU) to determine possible uses of errors and omissions on surveys of incoming and freshmen college students. The subjects of the first study were 1,927 individuals who had applied for admission to SLU and attended freshman orientation in the summer of 1989. Blanks and incorrect responses to a survey were taken as an independent variable, GOOF. A t-test found that the mean GOOF score of the respondents who enrolled that semester was significantly lower than the mean GOOF score of respondents who did not enroll. The second study involved a random sample of 1,540 new freshmen attending SLU in the fall of 1989. Two independent variables were the GOOF variable and one called RESPDNT for those subjects who did not return a mailed survey. Dependent variables were subsequent academic performance and fall-to-fall retention. Results indicated that for enrolled students the GOOF variable may not be an important one, but failure to respond (RESPDNT) to a survey may have meaning in terms of academic performance and retention. Included are appended instructions for creating the GOOF variable and six references. (JB)
Publication Type: Reports - Research; Speeches/Meeting Papers
Education Level: N/A
Authoring Institution: N/A