Mike, SPSS has a MISSING VALUES module that addresses your problem. It can analyze your missing values to ascertain whether there is a pattern, and it can also replace them with values predicted from a set of predictor values, not only similar questions but also variables such as sex, age, education and the like. You may also want to read the book by Roderick Little and Donald Rubun Statistical Analysis with Missing Data (J. Wiley & Sons, 1987).

Hector Maletta Universidad del Salvador Buenos Aires, Argentina

Michael A. Schoenfield wrote: > > I am working on a database of quantitative data obtained from a survey of > members of the American electorate on trust of government and I have a number > of cells with missing values. I am tempted to use the mean (average) of > answers to similar questions as a replacement for missing values. I am using > SPSS 7.0. > > I am hoping that there are alternative measures for estimating missing values. > Any suggestions. > > Mike S.