[Syllabus] [Schedule] [Homework] [Projects] [Bibliography] [Resources] [Directory] [Grads]
|
Weight |
Topic |
|
45% |
Study designs, measurement types, and which statistics to use with each. You may be given descriptions of studies and asked to describe their design and the kind of statistic you would use to analyze their results with. You may be asked to give an example of a particular kind of study design, including a research model and description of variables. You may be given an excerpt of data from a study and asked what kind of statistic you should use. Also important is an understanding of extraneous/confounding variables and how to deal with them, and a conceptual understanding of mediating and moderating variables, proximal vs. distal outcome variables, and independent vs. dependent variables (relative to a research model). |
|
10% |
Descriptive statistics. You should be able to determine the mean, median, mode, variance, and standard deviation for a data sample (when appropriate), and be able to construct a frequency distribution for it and be able to state whether it is unimodal or bimodal, symmetric or positively or negatively skewed. |
|
10% |
Definitions and concepts. You should be able to describe the following: Empirical vs. analytic research; scientific explanations; the scientific method; qualitative vs. quantative research methods (and examples of each); primary vs. secondary research literature; meta-analysis; the three ethical principles of human subjects research (from lecture); sampling; generalization of study results; ethnographic research; the typical use of different types of studies during the software development process; internal vs. external validity of a study; reliability vs. validity of a measure; system usability; coding manual (for behavioral/observational measures); inter-rater reliability; retrospective vs. prospective study; demographic measures; between subjects design; distribution of individuals vs. distribution of means; levels/treatments/conditions in a study. |
|
10% |
Hypothesis testing. You should be able to describe the basic logic of hypothesis testing, research vs. null hypotheses (and how they relate to the populations/population parameters in a study and whether the test is one vs. two-tailed), and significance levels. If provided output from R for Chi-squared goodness-of-fit, Chi-squared test for independence, Pearson correlation or t-test of independent means you should be able to state what the results mean in English (relative to the study hypotheses) and reformat the results into publication format. |
|
5% |
Sampling and generalization. You should be able to describe simple random, systematic, and stratified sampling, and what constitutes a representative vs. biased sample and how this affects the generalizeability of your study. |
|
5% |
You should be able to construct a scatter plot from sample data and be able to estimate (using the "eyeball" method) the Pearson correlation coefficient given a scatter plot. |
|
5% |
Measures. You should be able to describe the steps you would take in validating a measure, including test-retest reliability, internal consistency, and face, content, criterion, and construct (convergent and discriminant) validity, or identify when these are mentioned in the description of a measure. You will not be asked to calculate any of the figures, but you may be asked to interpret them. |
|
5% |
Survey measures. You should be able to describe and provide examples of: open-ended questions; restricted/closed-ended questions; rating scales (including Likert and semantic differential); and composite measures. If provided with a filled out survey index (composite measure) you should be able to compute the composite score. |
|
5% |
Power & significance.
Be able to describe the relationship between power, effect size, and Type I
and Type II errors, and the major study parameters that affect these. You
should be able to conduct a power analysis, given sufficient information, and
be able to compute effect size for two-group between-subjects and
within-subjects experiments. |
|
Weight |
Topic |
|
50% |
Study designs, measurement types, and which statistics to
use with each. You may be given descriptions of studies and asked to describe
their design and the kind of statistic you would use to analyze their results
with. You may be asked to give an example of a particular kind of study
design, including a research model and description of variables. You may be
given an excerpt of data from a study and asked what kind of statistic you
should use. Also important is an understanding of extraneous/confounding
variables and how to deal with them, and a conceptual understanding of
mediating and moderating variables, proximal vs. distal outcome variables,
and independent vs. dependent variables (relative to a research model). Includes
descriptive studies, correlational studies, demonstration studies, and
two-group, multi-group and multi-factor between-subjects
and two-group within-subjects experiments. Includes descriptive
statistics, chi-square goodness of fit, Pearson correlation, t-test for
independent means, t-test for dependent means,
one-way and multi-factor ANOVAs. |
|
10% |
Data screening. You should be able to describe how
to handle study data once it is acquired including: screening for outliers,
floor or ceiling effects or other violations of analysis method assumptions;
knowing how to apply data transformations and when they are appropriate;
checking for potential confounds (between-subjects) or order effects
(within-subjects); and how to do subgroup analysis. |
|
10% |
Definitions and concepts. You should be able to describe
the following: Empirical vs. analytic research; scientific explanations; the
scientific method; qualitative vs. quantative research methods (and examples
of each); primary vs. secondary research literature; meta-analysis; the three
ethical principles of human subjects research (from lecture); internal vs.
external validity of a study; reliability vs. validity of a measure;
retrospective vs. prospective study; demographic measures; between subjects
design vs. within-subjects design; levels/treatments/conditions
in a study; longitudinal vs. cross-sectional experimental design;
single-subject vs. group design; covariate; quasi-independent variable. |
|
10% |
Hypothesis testing. You should be able to describe the
basic logic of hypothesis testing, research vs. null hypotheses (and how they
relate to the populations/population parameters in a study and whether the test
is one vs. two-tailed), and significance levels. If provided output from R
for Chi-squared goodness-of-fit, Chi-square test for independence, Pearson
correlation, t-test of independent means, t-test of dependent means or
one-way or multi-factor ANOVAs you should be able to
state what the results mean in English (relative to the study hypotheses) and
reformat the results into publication format. |
|
5% |
Descriptive statistics. You should be able to determine the mean, median, mode, variance, and standard deviation for a data sample (when appropriate), and be able to construct a frequency distribution for it and be able to state whether it is unimodal or bimodal, symmetric or positively or negatively skewed. |
|
5% |
Study documentation. You should know the major parts
and structure of a study proposal and report/publication. For example, you
might be asked where your hypotheses should be stated, or what parts of a
sample proposal are missing. |
|
5% |
Power & significance. Be able to describe the relationship
between power, effect size, and Type I and Type II errors, and the major
study parameters that affect these. You should be able to conduct a power
analysis, given sufficient information, and be able to compute effect size
for two-group between-subjects and within-subjects experiments. |
|
5% |
Measure Validation. You should be able to describe the steps you would take in validating a measure, including test-retest reliability, internal consistency, and face, content, criterion, and construct (convergent and discriminant) validity, or identify when these are mentioned in the description of a measure. You will not be asked to calculate any of the figures, but you may be asked to interpret them. |