Research and Program Evaluation Flashcards
Idiographic
Single Subject
Nomothetic
Using groups of individuals to discover general principles
Kurt Lewin’s Action Research
Intended to improve the situation with local people/clients who will be better off at the end of research
Threats in Internal Validity - Were the dependent variables truly influenced by the experimental Independent variables or did other factors have an impact?
Maturation of Subjects - Psychological or physical changes
Mortality- Subject withdraw
Instruments used for measurement
Statistical Regression - The notion that extremely high or low scores would move towards mean if measure was utilized again.
External Validity
Can the experimental research results be generalized to the larger population? If not, EV is very low.
Parsimony (or Occam’s Razor) in Research
interpreting the results in the simplest way
Factor Analysis
Statistical procedure that uses the important or underlying “factors” in an attempt to summarize a lot of variables. Concerned with data reduction
Confounding
occurs when an undesirable variable which is not controlled by the researcher is introduced in the experiment
Quasi-Experiment
a type of research design that attempts to establish a cause-and-effect relationship by using other criteria other than randomization.
Minimum Subjects
Correlational Research - 30
Survey - 100
Organismic Variable
one that researcher cannot control yet exists (ex. height, weight, gender)
R.A Fisher
Pioneer of hypothesis testing.
Alternative Hypothesis (or Affirmative Hypothesis)
The IV has indeed caused change
probability (P) in research should be set beforehand and be…
.05 or lower. Two most popular- .05 (difference occur via change 5 out or 100 times of 5% of the time or 1 in 20 chance) and .01
Alpha (Type 1) Error vs. Beta (Type 2) Error
Alpha- Research rejects null hypothesis when it is true - probability equals level of significance
Beta- Accepting the null hypothesis when it is false - lowering significance level (P=.001) raises the chances of committing a Beta error.
t test
Simplistic form of the analysis of variance used to ascertain whether two samples means are significantly different. Computation (t value) must exceed the number cited in the t table in order to reject the null. More than two groups? Use ANOVA.
One-way vs. two-ways Analysis of Variance
One-way- testing one IV, even if it has more than one level
two-way- testing multiple IV,
Multiple DV? -MANOVA (multivariate analysis of variance)
Correlation Coefficient (r)
Makes a statement regarding the association of two variables and how a change in one is related to to the change in another. Ranges from -1.0 to 1.0. 0.0 = no relationship
Biserial Correlation
one variable is continuous (measure using interval scale)
one variable is dichotomous (two valued)
The Pearson Product-Moment Correlation r
used for interval or ratio data, most common correlation coefficient
Spearman rho correlation coefficient
Used for ordinal data
The Empirical Rule (68-95-99.7)
In a normal distribution, 68% of scores fall within +/-1 standard deviation of the mean; 95% within 2 SDs of the mean, and 99.7% SDs within 3 SDs of the mean
Mode
The point of maximum concentration
Factorial Design
Designing an experiment with more than 1 independent variables
Histogram
A distribution with class intervals that can be graphically displayed via a bar graph
x-axis
horizontal axis used to plot the IV. Also can be called the abscissa
y-axis
vertical axis used to plot the dependent variables or data. also can be called the ordinate
Scattergram
a pictorial diagram of two variables being correlated
z-scores = Standard deviation = standard score
Stanine scores
Standard Nine - nine intervals with 5 being the mean
4 basic measurement scales be complexity
N- Nominal - no true zero, no order, classifies/labels groups (nonparametric - non measurement)
O- Ordinal - orders variables but distance not always equal (nonparametric)
I - Interval - no absolute zero, but numbers scaled equal distance apart (Parametric, can + or - but but not / or x)
R- Ratio - interval scale with absolute zero. can use +,-,x,/
Hawthorne Effect
a tendency in some individuals to alter their behavior in response to their awareness of being observed.
The Rosenthal Effect
high expectations lead to improved performance in a given area.
