Statistics Speech

Why we rely too much on Statistics

Statistics? What does statistics have to do with psychology? Perhaps you have heard this question. It is a fairly common one and usually reflects the belief that psychologists are people who help others with emotional problems, mostly by talking with them. Why would such a person need to know about statistics? Of course, as this two-volume work shows, psychologists have many other interests and do many other things besides helping people with emotional problems. Nevertheless, even if psychologists restricted themselves to this one endeavor, statistics would be a necessary component of their field because the particular techniques that clinical and counseling psychologists use were selected after being tested and compared with other techniques in experiments that were are analyzed with statistics.

Of course, because psychologists are interested in memory, perception, development, social relations, and many other topics as well as emotional problems, they design and conduct experiments to answer all kinds of questions.

Statistics are mathematical techniques used to describe and draw inferences from quantitative data. Statistics are commonly divided into two branches: descriptive and inferential. Descriptive statistics are used to describe, summarize, and represent more concisely a set of data. Common descriptive statistics include frequency distributions, percentiles, the mean, and the range. Inferential statistics involve procedures for drawing inferences that go beyond the data set: conventionally, inferences about a large group (i.e., a population) based on observations of a smaller sample. Statistical inference could be used, for example, to estimate the relationship between variables (e.g., the correlation between income and amount of services received), to assess whether two groups differ (e.g., does the treatment group have better outcomes than the control group?), or to judge the fit of a complex model (e.g., how well does the program theory fit the obtained data?).

Statistics is divided into two categories: descriptive and inferential. The objective of descriptive statistics is to describe or summarize the properties of data that a researcher has collected. Inferential statistics is for inferring from a sample to a population.

In general, the goals of qualitative research (i.e., to identify key trends or patterns and transferable findings and to study in-depth phenomena within small populations) are quite different from the goals of quantitative research (i.e., to discover generalizable results and to study across broad populations); therefore, the use of statistics is not the same in the two approaches. As with descriptive statistics, inferential statistics allow researchers using a multi or mixed method approach to provide another context, a richer picture or enhanced representation, in which to examine the phenomenon of interest. The inclusion of quantitative data can also enhance legitimacy (i.e., validity, credibility, trustworthiness, or transferability), although this may not be appropriate for some qualitative studies. Not all researchers agree with the mixed method approach.

Qualitative and quantitative researchers use three inference modes: abduction, induction, and deduction. Abduction occurs when one studies the facts, hunches, and data and develops general propositions from specific observations to devise a working hypothesis. Inductive reasoning occurs when one tests the working ...