Enlightenment effects—occurs when psychological research is common people knowledge, and knowing theory will make people change themselves.
- Social Psychology is primarily a historical inquiry because it is temporally, contextually and historically bound.
- Argues psych is an open system so there are too many rival explanations.
Gergen: SCIENCE: “Social Psych as History”
Refutes Gergen- Psych can be compared to natural hard sciences
- People have free will but they also are predominantly predictable.
- Explanation of human behavior is possible.
- Believes Gergens’ effects are real but can be avoided with methodological creativity.
Schlenker: SCIENCE: “Social Psych as Science
Near tautology is like a universal truth, can’t be proved empirically, can’t be disproved
- Too much research seeks to prove near tautologies; waste of time in research.
- Only three instances when they’re useful: 1) Trying to change a stubborn dependent variable 2) when small changes in IV will change your DV dramatically 3) When multiple IVs can influence the DV.
Wallach & Wallach: TAUTOLOGIES: “Are hypotheses in social psychology empirical?”
Authorship is very important for career advancement for both faculty and students, which can cause dilemmas.
- Two dilemmas: When faculty take authorship credit that is actually earned by students and when students are granted undeserved authorship
- Three ethical principles: beneficence (supervisors should help students further their career by giving authorship only when deserved), justice (treat fairly and give each contributor what is deserved), and parentalism (inappropriate for supervisors to use position of power to make decisions).
Fine: AUTHORSHIP: “Reflections on determining authorship credit and order”
A discussion about authorship between all individuals should start from the inception of the project and remain a dynamic process throughout.
- Authorship requires professional contribution, not just time put into a project.
- An authorship-determination worksheet can be useful because it provides a quantitative example for determining authorship and clarifies expected roles.
APA: AUTHORSHIP: “Student’s Guide to Determining Authorship Credit and Order”
Researchers have an ethical imperative to consider the cost and utility of conducting research as well as that of not conducting research.
- All else equal, research of higher quality is likely to be more ethical because resources are better invested.
- Since the cost of research can be great, researchers should take advantage of a variety of data analyses to make best use of the information available.
Rosenthal: ETHICS: “Science and Ethics in conducting, analyzing, and reporting”
“Researcher degrees of freedom” are the many ambiguous decisions researchers must make and may increase the probability of false positive results. Four common decisions are: choosing among dependent variables, choosing sample size, using covariates, and reporting subsets of experimental conditions.
- False positives, when published, can become persistent and lead others to pursue fruitless research.
- Simmons proposes disclosure requirements for authors and guidelines for reviewers in order to find a middle ground between “too far” and “not far enough” in service of providing publications that are accurate and transparent.
Simmons: “False-positive psychology”
The prevalence of questionable research practices (QRPs) puts the integrity of research results in jeopardy and is higher than previously thought.
- While some researchers viewed their own QRPs as defensible, they viewed QRPs as indefensible overall.
- When studies are often published using QRPs, it pressures other researchers to pursue results that can only be obtained by also using QRPs and negatively impacts research field overall, creating a “race to the bottom” (wasting resources and damaging reputations).
John: “Measuring the prevalence of questionable research practices”
External Validity is only important when the goal is generalizability.
2 and 3. Four types of experiments that don’t require external validity:
- Exploring what “can” vs. what “does” happen
- Specifying something that “ought” to happen in lab
- When demonstrating an effect happening in unnatural conditions
- When the researcher purposefully wants to produce conditions that have no real world meaning.
Mook: “In defense of external validity”
Inductive approach to external validity: (specific to general) This means trying to take what is known from all studies and then making broad generalizations.
- Deductive approach to external validity: (general to specific/ill-advised) Establishing standards for the generalizability of an experiment depends on certain “essential features” that straddle both the lab and the field. These “certain features” are impossible to know in advance, however, and require inductive logic to determine.
- Experimental realism is more important than mundane realism
Locke: “Generalizing from lab to field: Ecological validity or abstraction of essential elements”
Demand characteristics are an experimental artifact caused by social and environmental cues. These are anything in an experiment that is not meant to be the manipulation and is affecting the outcome (Ex. Panic button, lab coat).
2. Participants subject to demand characteristics form an interpretation of the experiment’s purpose and unconsciously change their behavior accordingly.
Orne: DEMAND CHARACTERISTICS: “The psychology of the Panic Button”
The study reexamined extraneous variables and confounds that produced effects on morale, supervision, and group influence at Hawthorne.
- Old Hawthorne Effect: If there are any changes in the environment, there will be changes in productivity. Improper methods, analyses, and reporting by the original experimenters led to these invalid results, which persisted until reexamination of the study.
- New Hawthorne Effect: Changes in productivity are caused by combination of informational feedback and financial rewards, not simply changes in the environment.
Parsons: CONSTRUCT VALIDITY: “What happened at Hawthorne”
We have moved away from a justificationist view of science towards a nonjustificationist view, which is the stance that each component of theory tested must be characterized as an ongoing process that is open to criticism and evaluation. A theory is never fully proved or disproved.
- Smith proposes a 5-step model to establish construct validity.
- Generalizability theory (GT) aims to estimate different sources of variance in test scores (e.g. interviewers, raters, items) and thus facilitates calculating one overall reliability estimate. GT provides an increased capacity for controlling measurement error.
Smith: “On construct validity”
Confirmation bias is the tendency for judgments based on new data to be overly consistent with old hypotheses. It delays finding support for beliefs other than those presently held.
- The example of research on the sleeper effect is an example used to demonstrate how confirmation bias can steer research in the wrong direction. A sleeper effect is an effect that takes some time to become apparent. Researching the conditions under which an effect occurs as opposed to whether or not it exists would have led to knowledge on the phenomenon much sooner.
- There are two alternative methods to avoiding theory confirmation: a) condition-seeking and b) design approach.
Greenwald: CONFIRMATION BIAS- Sleeper Effect: “Under what conditions does theory obstruct research progress?”
Traditionally, basic and applied research have been dichotomized, with the former generally treated as superior to the latter.
- There is now a renewed understanding that both are equally important. Basic researchers provide theoretical insights that applied researchers need in order to address social problems. On the other hand, applied researchers offer basic researchers insight into which concepts are practically important and merit further study.
- Zigler argues that changes must be made in professional training and expectations to prepare graduate students for both types of work.
Zigler: “A place of value for applied and policy studies”
