Definitions• Theory o Systematic way of organizing/explaining observations• Hypothesis o Tentative belief/prediction about the relationship between two or more variables• Variable o Something that can change or vary from one situation or person to another (IQ score, age, sex, education, height)• Operational definition • Clear “label” of how a variable is measured
Definitions• 2 types of variables o Independent variable • Potential influences on behavior under study • Age, gender, advertisement, SES, stress o Dependent variable • “Outcome variable” or “outcome factor” • A behavior, action, attitude, performance, score on tests/surveys
Definitions• Variable measurement o Continuous (degree of optimism, rate of recovery, shyness) vs. o Categorical (gender, species, history of cancer)
Sampling• Population = a larger group from which to draw a sample• Sample = a subgroup representative of the population• Participants = the individuals participating in the study• Generalizability = the applicability of findings to the entire population of interest• Random Sampling = the sample should be similar to the larger group as every person in the study has an equal chance of being placed into the treatment condition
Measurement• Reliability o A measure’s ability to produce consistent results o “Internal consistency” • Degree participants’ scores on a measure resemble their scores on the same measure when administered later
Measurement• Validity o A measure’s ability to assess the variable it is supposed to assess o Are you measuring what you think you are measuring? o Note: Prior validity is not an indication of present validity. WHY?
Measurement• General Sources of Error o Random = due to chance events (i.e. misrecordings) o Coverage = failure to include relevant people in a sample (i.e. random digit dialing) o Nonresponse = refusal to respond or absent
Research Methods• Descriptive (Case) studies • Poor sample size (N = 1) • Selection bias (nonrandom sample) • Observation bias (is the therapist seeing what he/she wants to see?) • Causal factors unidentified• Correlational Studies • Do not specify cause/effect relationships • Can help identify naturally occurring relationships• Experimentation and Surveys • Accuracy of self-report? • Low response rates • Social desirability bias = try to make yourself look better than you are • Demand characteristics = say what you think the survey wants to hear • Issues of generalizability outside the lab?
Surveys• What do surveys measure? o Attitudes o Behaviors o Beliefs o Opinions o Preferences o Satisfaction o Demographics o Orientations (religious, sexual, political, etc.)
Surveys• Principles for writing survey questions: Choose simple over specialized words (“apartheid?”) Use complete sentences to ask questions Avoid vague language and/or cultural slang Provide appropriate time referents Limit the number of personal questions Keep instructions brief! Ask only one question at a time Start with the most salient questions in case of drop-out When possible, group similar response sets together Use dark print for questions + light print for answer choices
Surveys List answers vertically instead of horizontally Avoid double-barreled questions Keep font simple and “clean” Limit or eliminate skip patterns Do not use emotionally charged words (“hate”, “fantastic”) Do not use leading questions Use a scale that covers the full range of response possibilities Conduct focus groups in pre-testing phase (qualitative output)
Experiments• There are two types of information we want: 1) Description/Prediction (“what?) What are people doing? Thinking? What is happening? What do people do in a certain situation? 2) Explanation/Understanding/Control (“why?”) What causes people to do or think like they do? Why is it happening? How can manipulating something (attitudes, situations) change people’s behavior?
Experiments• Correlational relationships o Can indicate if two variables are related but do not indicate which variable influence which o If variable A and variable B are correlated, we cant know which variable influences which. Why not? • does A cause B? • does B cause A? • does C (some other factor) cause both A and B?
Experiments• Correlation coefficients range from – 1.00 to + 1.00 o Positive correlation: 0 to +1.00 o Negative correlation: 0 to –1.00• The sign of the correlation (+/–) indicates the kind of relationship (but not the strength of the relationship) o Positive correlation: the more cars on the road, the more smog there is in the air o Negative correlation: the fewer days you miss in my class, the higher your final grade
Evaluation of Researcho Does the theoretical framework make sense?o Is the sample appropriate?o Are the measures and procedures adequate?o Are data conclusive?o Are data interpreted correctly?o Does the study add anything meaningful?o Is the study ethical?
Ethical Guidelines• Human subjects research Confidentiality and anonymity of participants Informed consent Be clear participation is voluntary Limited to no deception allowed Lack of coercion Minimize harmful consequences
A particular slide catching your eye?
Clipping is a handy way to collect important slides you want to go back to later.