|
|
|
|
|
Inferential Statistics Key Terms or SPSS Outputs to Understand
Module Objectives: At the end of this module the student should be able to: 1. Explain the formulation of Hypothesis Testing in Health Science and Medical Research 2. Assess and explain the Meaning of Different Levels of Significance in Health Science Research 3. Explain and justify inferential statistical procedures used to test or describe specific data and discuss the appropriateness of it for answering the study hypotheses. 4. Explain how to properly use SPSS to perform a t-Test to analyze the differences between the means of two groups. The student will also learn how to interpret the SPSS output for various statistical tests. This course is based on the assumption that students accepted into this course already have had the appropriate statistics and research methods courses in their earlier studies. However, it may have been a few years since some of the students have had an opportunity to review some of the material covered in this course. If you need to refresh your memory concerning statistics/biostatistics please review some of the sources presented on the Background page for this module as well as Module One Background. It is recommended that you bookmark those reference sources you find as valuable. Also be sure to be very conversant with all Key Terms presented in each Module of the course. Introduction One of the cornerstones of research is hypothesis testing. A clearly written hypothesis statement gives direction to a study in much the same way that a well defined and written research question guides the researcher. However, there are significant differences between them. First, the hypothesis statement is a form of a decision point in the research process. It defines the point that the research study's null hypothesis is either rejected or accepted and, therefore, the acceptance or rejection of the hypothesis. In experimental research determining the level of significance of research findings and the decision point for acceptance of a study's findings is one of the most important aspects of designing a study. Secondly, hypotheses are statements of relationships found in populations. This differs from research questions which frequently describe the relationships found in a population sample. The value of the hypothesis is derived from whether or not it is testable in the real world. In the health sciences and medicine, a hypothesis's value is also derived by its ability to provide a decision point (acceptance or rejection) which is clinically significant. This point is very important since a hypothesis can provide a researcher with a finding which is clinically significant but statistically insignificant. There are also ethical issues to consider when a treatment or a life saving procedure is rejected solely on the basis of it being found to be statistically insignificant while practicing physicians find it to be clinically significant. In this module we will examine the importance of inferential statistical procedures which utilizes research data sampled from a study population to make inferences about that particular study population. (It is highly recommended that you seek a detailed description of inferential statistics. See this site and search for Inferential Statistics.) Inferential statistics generally come from the family of statistical models known as the General Linear Model. (Search for General Linear Model.) The statistical methods used frequently in research are the t-test, ANOVA, ANCOVA, MANOVA, MANCOVA, regression analysis, factor analysis, and cluster analysis among others. For the purposes of this course it is important to generally the types of research problems that each method could be applied to and the problems associated with each method. This topic is especially important for students conducting experimental or quasi-experimental research. Important Concepts/Key Terms Range Effects: The sampling procedure or measuring instrument can result in a study population which is restricted in range when compared to the general population. For instance, the range in heights of basketball players is much more restricted than that of the general population. In other words, when you measure a restricted group (5th graders compared to all grades, IQ of graduates students compared to the general population) the range of values will be much more restricted than that for the general population. Now, imagine how this fact could effect the results of a various research studies! Outliers: Many studies may be influenced by outliers. Depending of where the outliers are found in a data set they can influence the value of the sample mean and a correlation coefficient. Nonlinearity of Regression: This occurs when the assumption of a linear relationship between the variables doesn't actually exist. The result is a low correlation coefficient and an underestimate of the true relationship between the variables. Measurement Reliability: The reliability of the measurement instrument must be sufficiently high enough not to have an effect on the actual data analysis. Subjective Judgment of Persons: Expectations and intuition of people lead to sometimes false correlations or conclusions. An example is the belief that tall people always have tall children and obese parents have obese children. Subjective judgments sometimes influence some studies.
SPSS -- Working with SPSS -- Testing the Difference Between Two Means -- Inferential Statistics |
|
|
