Exam 1 Flashcards
What is quantitative design?
Numerical Data- can be analyzed statistically
What is Qualitative design?
non-numerical data- naturalistic observations etc.
What is the scale of measurement?
N- Nominal
O- Ordinal
I- Interval
R- Ratio
What is nominal scale of measurement?
Categorical
Measures only differences
No order
What is ordinal scale of measurement?
measures (rank order) magnitude
preserves position- rank your favorite ice-cream
What is interval scale of measurement?
Has equal intervals
temperature, no true 0 point
What is ratio scale of measurement?
equal intervals and a true 0
can apply multiplication and devision
height, weight, kelvin, etc.
What are the four types of quantitative design?
description, association, true experiment, quasi experiment
What are descriptive design?
1) Frequency
measures one or more dependent variables.
descriptive stats only
What are associations?
relational research
measures at least 2 dependent variables and how they are related
- Does not imply causation
What is a true experiment?
manipulate at least 1 independent variable
measures at least 1 dependent variable
- uses inferential statistics to make causal inferences.
quasi experiments
oftern used when parts cannot be randomly selected or assigned to condition- gender, age, IQ
- teat the subject like a IV
What is frequency distribution?
a description of how frequently each score appears in the data.
What is a histogram?
is used to describe a frequency distribution for continuous data
frequency on the Y- equal size bins on the x
Whats the job of descriptive statistics?
used to describe a distribution
- Where is the center and how wide is the distribution?
What are the measures of central tendency?
1, Mean- can only be used for inter or ratio scales
- Median- OIR
- Mode- NOIR
What are the measures of dispersion?
- range
- variance/ SD- average squared deviation of scores from the mean
variance and SD can only be used for interval and ration scales.
What happens to variance as you add more scores?
it reduces
What is a normal distribution?
- symmetrical, bell shaped curve
2. all measure sof central tendency are the same- Mean= median= mode
What is a skewed distribution?
long tail on one end- mean, median, mode not equal.
which measure of CT is most influenced by extreme scores?
mean
What are some solutions for skewed data?
- try to reduce variability in data- collect more data or practice trials
- impose time limits for RT
- use the median as your measure of central tendency.
- Use an a priori method to trim the mean
What is the defining characteristic of an association?
- measures the degree of relationship between two or more measured variables.
what do associations allow us to do?
make predictions- class attendance is a predictor variable for final grade. - predictions are not perfect
what statistical analysis do you use for associations?
- Chi-squared- NOIR
- spearman rank correlation coefficient- OIR
- Pearson correlation coefficient (r)- IR
What is a qui squared test?
- Determines whether the two variables are independent.
- p value indicates probability that the relationship is due to chance.
Contingency Coefficient
- converts chi2 value into a coefficient showing the degree of relationship
- 0 no relationship, 1 strong relationship
- significance depends on strength of relationship and # of observations.
What do the direction of correlations mean?
- Positive Correlation- as a increases b increases and vice versa
- Negative Correlation- as A increases, B decreases
What is the scale of a Correlation?
-1.0-1.0
What is a dependent variable?
- its what is measured by the experimenter.
- can be subjective or behavior/physiological
- many exp have more than 1 DV
What is the speed accuracy tradeoff?
occurs when participants adopt different strategies in different conditions.
- therefore, it is customary to record and report both
What are the two types of independent variables
- Manipulated IVs-
2. Non- manipulated/ quasi
What are some characteristics of IVs?
- all have a name and at least two defined levels
2. sometimes one of the levels is called the control condition
What is a hypothesis?
describes what the predicted relationship is between the IV and DV
- expect a manipulation in IV to cause a change in the DV
What is the peer review process?
the process by which research papers are reviewed by respected scientist in the same or similar field prior to their publication in a journal.
- reviewers provide feedback
- paper is then accepted or rejected
What happens is the Journal has been rejected?
- the author may be invited to revise and resubmit the paper
- raises the overall quality of published journal articles.
What is a primary source?
- a publication in which the author presents his or her own research.
What is a secondary source?
- author does not present his or her own research, but instead reviews previously published literature on a particular topic.
What is the advantage of psych info?
- most peer-reviewed publication.
What is the difference between an experiment and a correlation in an abstract?
- TE: indicates IV and DV; Key terms (effects/influence, manipulate, condition, assigned”
- Correlation: multiples measurements, no mention of manipulated variables; Key terms: X predicted Y, X was associated with Y, X was related to Y
What is internal validity?
- degree to which a change in the dC is attributable to manipulation of IV
- extent to which an expt. is free from confounding variables
What are the threats to internal validity?
- extraneous variables that covary with IV levels and provide an alternative explanation for the results (always a problem for quasi experiments
How do you remove confounds?
- try to control, constrain, or randomize extraneous variables
What is external validity?
- degree to which a research finding can be generalized to other populations, settings and times.
What are threats to external validity?
- too many control variables (always a problem for laboratory experiments.
How do you improve external validity?
- try to randomize some variables
- repeat the experiment with other participants, in other settings, with other stimuli.
What are demand characteristics?
- clues that influence the participants behavior in your experiments
- 2 types: participant bias and experimenter bias
What is participant bias?
- when behavior of participants in influenced by their beliefs about how they are supposed to behave in a study.
What is experimenter bias?
- when an experimenters expectation about an experiment influence the outcome?
How do you minimize demand characteristics?
- automation- computerized instructions
- post experiment questioning
- double blind experimentation
- use deception
What are typical confounds in BS design?
- anytime there is a difference between your groups other than the variable of interest (always a problem for quasi experiments)
What are typical confounds in WS design?
- anytime there is a difference between you conditions other than variable of interest