laurie@cybavax.UUCP (Laurie Moseley) (04/24/86)
MICRO-COMPUTER STATISTICS PACKAGES I am looking for a relatively simple statistics/ data analysis package for use on an IBM-PC or AT. The statistics do not need to be sophisticated. What we need de minimis are: 1. Data-manipulation: Facilities for grouping (like RECODE in SPSS) Facilities for constructing variables (like COMPUTE in SPSS) Preferable (but not essential) data selection (like SELECT IF) Preferable (but not essential) logarithmic transformation Preferable - some simple graphical display e.g. scattergrams, or possible linking to a graphics package. 2. Data presentation Cross-tabulation with percentages Preferable - 3/4 way cross-tabulation (but not essential) Mean, median, SD Equivalent of SPSS BREAKDOWN Some form of labelling 3. Stats (All desirable but not essential) Chi-square /Fisher exact Spearman or Kendall non-parametric correlation Kendall coefficient of concordance Some simple linear regression There are plenty of packages on the market. However, I have no idea how well they work in practice. I would be grateful if anyone who has had either very positive or very negative experiences with any such packages could share that experience with me. Perhaps I should note that the package is intended for desk-top analysis and data-exploration by some social scientists (and I mean social scientists not woolly theoreticians !). Of course, if there is any public domain material, we would be overjoyed to hear about it. With thanks Laurie Moseley University College, Swansea -- If you have skill, you have power. If you have power, you need responsibility. Never demonstrate one without the other.