| 1. Refer
to my web pages for valuable assets. You can prepare for Ed
711 and Ed 710 content using
web sites that I have used in class. These make excellent
refresher
materials if you haven't used these web pages before. One link
that
you will want to use all the time is entitled Navigate
with Bill Trochim. The links for Ed710 take you to
web
pages that provide terrific background on multivariate
statistics.
These pages connect you to other sites throughout the internet that
link
to even more helpful resources of every stripe. A New
View of Statistics is an indexed site. It may be the best yet for
statistics
review
2. Refer
to my web page for online links to recent materials for Ed
714 Qualitative Research Methods in Education. The
syllabus
is essentially the study guide for the course. 3. Guidelines for the development of objective tests can be found on my web page as well. Go the link Guidelines for Objective Tests. 4. A set of guidelines for the development and validation of performance assessment tools are helpful as well. Go to the link Performance Assessment Guidelines. 5. Find additional information regarding the design of surveys and questionnaires at The STATPAC website 6.
Study
guide topics are listed below. Sampling What are the various methods that are classified as scientific? How is each type of sample drawn? When is each used? What are the advantages of each, as they apply to sampling error? What are the limitations of each approach? How is sampling error calculated? What is the relationship between sampling error and generalizability, between sampling error and power? What is the benefit of large sample sizes? When and how are large samples misleading? What are weighted samples? How are they drawn? When is each used? What are the advantages of each? What are their limitations with respect to sampling error and sample reliability? What
is the difference between generalizability and confidence? Instrumentation What are the various types of statistical reliability? Their statistical coefficients, purposes and questions they answer? What is meant by errors of prediction? What is the relationship between accuracy and generalizablity of scores? What are the various types of statistical validity. Their coefficients, purposes and questions they answer? What is meant by errors of estimation? What is meant by each term and how is each used? Content validity, Logical validity, Conclusion validity? Why is test validity different from experimental validity? What is test bias? What are sources of bias? List and explain various approaches to the statistical demonstration of test bias? What are the basic steps in the construction of an objective test? What is power? What
are item scaling techniques? How are they calculated? When is each type
employed and what questions are answered by each approach? Research design What are Type I and Type II errors? How are they different? What do they tell one about the soundness of statistical inference? What is power? How
is power calculated? What are the ways in which research improve the
power
of a research design? Define nonexperimental, experimental and pseudoexperimental designs. Give examples of each. Match designs to illustrative research plans that capture the important features of each category. What are the most common threats to the internal validity of designs? How is internal validity different from internal reliability? What are the most common threats to the external validity of designs? Create examples of each type of threat to internal and external validity. Explain what you would do to control these problems. How
is the alpha level of a test different from the probability level of a
statistic? Statistical design What
is inference testing? List and describe the most commonly used types of
parametric and nonparametric What is the critical value of a statistic? Give examples, using statistical tests as your framework (e.g., the t-ratio and the t test). What is meant by the term, general linear model? What are commonly used multivariate designs? When are they used? What types of questions are addressed by each design respectively? What statistics are generated by each model? How are they interpreted? What assumptions of the parametric normal distribution must be met by parametric analyses? What assumptions of the multinormal parametric distribution are added to those previously mentioned? What,
if any, unique assumptions are made by analysis of covariance, repeated
measures analysis of variance, MANOVA, or factor analysis? Can you
think
of any other unique issues that apply to multivariate tests? Page
created November 1999. Modified May 2003
|