T HE purposes of this paper are: (1) to discuss briefly the relationship of statistics to quantitative scientific research, (2) to outline some of the uses of statistics in farm management research, and (3) to list and discuss some of the arguments given for not using modern statistical methods in collecting and interpreting farm management data. Many writers, including some who have written textbooks on statistics, have confused the discussion of the use of statistical methods by classifying statistics as a method of research on the level with experimentation and other methods. On the contrary statistics is a tool which is used in all quantitative research. It is true, of course, that the statistical methods used may be simple or complex; they may be efficient or inefficient and they may give biassed or unbiassed estimates, tests, or predictions. At the minimum, data are usually summarized by averages of one sort or another. These averages are statistics and are estimates of some population values or parameters. Since the data are usually based on samples, these estimates have a sampling error. In other words, the estimates would vary from sample to sample. Appropriate statistical methods for analyzing farm management data cannot be discussed intelligently without considering the objectives of a study, the specific hypotheses to be tested and the sources of data. Statistics can make an important contribution to the success of a research study in the planning stage. A majority of the studies that end in disappointment failed in their conception and design. Other studies fail because of improper analysis and interpretation of the data obtained, but these are relatively few in number compared to the former. The main steps in scientific quantitative research might be classified as follows:
Read full abstract