Research Methods Flashcards
Type of research that contributes to a theory
Basic Research
Type of research that helps answer the question of whether or not a theory helps solve real world problems
Applied Research
Whether or not a study can be replicated to get the same results
Reliability
Idea that a findings should be explained in the simplest way possible.
Parsimony
When unknown influences on the results make the study inaccurate
Confounding
The variable that an experimenter manipulates
Independent Variable
The outcome of the experiment (type of variable)
Dependent Variable
This variable is always some form of human response
Dependent Variable
The group that does not receive the IV
Control Group
The group that does receive the IV
Experimental Group
Question you’re trying to find out in an experiment
What is the impact of IV on DV?
Avoids sampling bias and keeps the researcher honest
Random Sampling
Random Sampling AND ______ are necessary.
Random Assignment to Experimental and Control Groups
When subjects cannot be randomly assigned or the experiments lacks in the control group this is called…
Quasi-Experimental Research
All intact group studies are considered ______
Quasi-experimental
When the people from subgroups are selected this is called
Stratified Sampling
When people are selected from subgroups to mimic the ratios of the general population
Proportional stratified sampling
When you use naturally existing groups and randomly select participants from the cluster
Cluster Sampling
“nth” sampling or “k-ith” sampling when, for example, every 10 person is chosen for the study
Systematic Sampling
The ______ of the sample is more important than the procedure used to pick the subjects
representativeness
Judgment sample, convenience sampling, and quota sampling are all…
Non-probability Methods
Sampling that is thought to be representative of the population by the researcher
Judgment Sampling
An existing group is used with no random sampling
Convenience Sampling
Pre-specified characteristics are used that mimic what is likely to be representative of the population in the future.
Quota Sample
The hunch a researcher has
Experimental Hypothesis
The experimental hypothesis is written as
Capital H with a small 1 next to it. It could also be written with a tiny a.
the short version of additional hypotheses
H2, H3, etc.
When the researcher predicts no significant difference between the control or experimental groups, this is called
Null Hypothesis
Null Hypothesis can be written as
H with a tiny 0
The tense used to write a hypothesis in the modern form
Present Tense
Describes how likely it is that the difference between control and experimental groups occurred by chance
Level of Significance (AKA Confidence Level or Alpha Level)
Short form to describe that there’s a 5% chance that difference between groups occurred by chance
p=.05
Type of error in which you reject null when it is true- REBT helped with alcoholism, but really it didn’t
Type 1 Error
Type of error in which you accept null when it is false- REBT did not help alcoholics, but it did
Type 2 Error
Tells if the experiment measures what it is supposed to measure
Validity
Tells if the results can be generalized to other groups in the real world
External Validity
When people perform better because they know they are being observed
Hawthorne Effect
Tells if the study forms a valid cause/effect
Internal Validity
Instrumentation, measurement methods, observer judgment, maturation, statistical regression, selection of groups, attrition, and demoralization are all threats to
Internal Validity
When time impacts the results this is called
Maturation
When re-administering a test affects the results it is called
Statistical Regression
Two terms used to talk about whether or not test subjects stay in the study for the duration
Attrition or Experimental Mortality
To test the significance of a study, this tests the hypothesis between two normally distributed samples
T-Test (appropriate for 30 or more subjects)
Conducting a T-Test on two separate occasions
Dependent or Correlated T-Test
Needs to be used to analyze variants if there are more than two groups in an experiment
an ANOVA
A ratio of two variances (how far the data are scattered from the mean)
F Statistic
F-Tests
Named after Fisher
Needed if there is more than one Dependent Variable
Multi-Variant Analysis of Variants
MANOVA
Needing to adjust groups so that a variable that might correlate with the DV will not throw off the study - for example if you find out one of the groups has read a book on REBT before the REBT study
ANCOVA - allows the researcher to statistically remove the effect of extraneous variables
Type of research that looks at whether a relationship between two variables exist, and if so, what is the magnitude and direction
Correlational Research
Correlation of choice is called
Pearson Product Moment Correlation
The range of correlations is
-1 through +1
When the correlation means that one goes up and the other goes down
Negative Correlation
When the correlation means both go either up or down
Positive Correlation
The mean, median, and mode all fall in the middle of this
Bell Curve or Gauzean Curve
Mean Median and Mode are all
Measures of Central Tendency
In a bell curve, when there are extreme values and the curve leans to one side or the other
Skewed Population
Bar Graphs are all called
Histograms
A measure of variability that shows the difference between highest and lowest values
Range
The measure of dispersion is called
Variance
Square root of the variants, useful in discussing spread of scores
Standard Deviation (SD)
This is not experimental and just describes a group (i.e. average, range, etc.)
Descriptive Statistics
Uses numbers to classify, but is qualitative, not quantitative
Nominal (Scale of Measurement)
Describes variables that can be rank ordered (high, medium, low)
Ordinal Scale
Scaled at equal distances, but no zero point (you can receive a zero on a test, but you don’t have zero knowledge)
Interval Scale
Does have a true zero point and each number is measured from zero (weight/height)
Ratio Scale
The simplest approach to research
Survey
When using a questionnaire, 30-50% return rate is typical, but you can’t make generalizations under
75% return
looks at overall dynamics of culture or situation rather than a single factor- holistic and inductive (case studies)
Ethnographic Research
generalize based on observations
inductive reasoning
when hypotheses derived from general principles
Deductive Reasoning
Rate an individual on one characteristic but really you’re influenced by another
Halo Effect
Experimenters expectations influence subjects
Rosenthal Effect
In testing or experimentation - refers to typical person who was tested
Norms
The letter N is used to represent
the number of subjects in a study
Untransformed scores
Raw scores
follow same people over a period of time (trend studies)
Longitudinal Research
just says there will be a difference but doesn’t say how
Non-directional Hypothesis
asserts how things will be different
Directional Hypothesis
The probability of committing a type 1 error equals
The significance level set for the experiment.
If the significance level is .05, you are accepting a 5% chance that you are wrong when you reject the null hypothesis.
What is stanine score?
Standard 9- divides into 9 and 5 is the mean. There is also T-scale where 50 is the mean.
