Home Page

Previous Chapter 9 Collecting Research Data through Observation and Content Analysis

#p288-9: #1 (Explain the relationship between descriptive research and causal-comparative research)

The purpose of Descriptive Research

· Type of quantitative research that involves making careful descriptions of an educational phenomenon.

· Determining what is?

p290

Description of a Sample at One Point in Time

· Descriptive research often involves reporting the characteristics at one point in time or a snap shot.

p296

· Causal-Comparative research is a nonexperimental investigation where you seek to identify cause effect relestionships by forming groups of individuals where the independent variable is present or absent and then determining if whether the groups differ on the dependent variable.

· Observe natural causes and not manipulated causes.

#6(Describe three possible interpretations of a relationship between variable A and variable B that is discovered through causal-comparative research)

P298 –

· major advantage: allow us to study cause and effect relationships under conditions where experimental manipulation is difficult or impossible.

· The major disadvantage of causal-comparative research is that determining causal patterns with any degree of certainty is difficult. An observed relationship between variables a and b can mean

1. A causes B

2. B causes A

3. or C causes both A and B

#8(Define suitable comparison groups, given an initially defined group in which a particular characteristic is present)

2. Selecting a defined group is the second step in causal-comparative research. You should select a group that possesses the characteristics one wishes to study.

· The definition should be precise so that the results of the study can be interpreted meaningfully.

P301

3. Selecting comparison groups is the third step in causal-comparative research

· Select a group that does not have the characteristic that you wish to study.

P302

· Matching is used to equate 2 or more groups on one or more extraneous variables so that these variables do not confound the study of the variables of interest.

· Extreme-group method : selecting comparison groups that are at the 2 extremes of a score distribution on one variable.

#9(Explain how the t test is used in causal-comparative research)

P303

Data Analysis of Causal comparative studies

· 1st step is to conduct an exploratory data analysis and compute descriptive statistics for each comparison group in the study.

· Next step is to do a test of statistical significance.

P304

· Statistical Analysis: The t-test for differences between Means (used to test the significance of the difference)

o T-test for small samples under 29; but are generally used

o Z distribution for large samples above 30

o T-test in causal comparative research depends on 3 assumptions about the obtained scores:

1. the scores form an interval or ratio scale of measurement.

2. the scores in the populations under study are normally distributed.

3. the scores variances for the populations under study are equal.

#11(Interpret the t value resulting from a t test for a single mean, and describe how this test is used in causal-comparative research)__[Also see p. 138-140 for review of significance tests]__

P305

· As you increase the number of t test you increase the chance of a Type I error.

· Analysis of Variance is a statistical procedure that compares the amount of between group variance in individuals’ scores with the amount of in group variance.

p306

In most causal-comparative studies: researchers compare the mean scores of 2 samples to determine whether they are significantly different from each other.

P308

· Analysis of Covariance: the need to determine a difference between 2 groups of a particular variable can be explained by another difference that exist between the 2 groups.

· Is used to control for initial differences between groups.

· Useful in casual-comparative studies.

P309

· Multivariate analysis of variance: used to determine whether groups differ on more than on dependant variable. (presumed effect) Independent variable is the presumed cause.

P313

· Parametric statistics: tests of statistical significance that are based upon certain assumptions about populations parameters.

Home Page

Next Chapter 11 Nonexperimental Research: Correlational Designs

Planning Page for this Wiki Check here to see who is doing what!