Presentation is loading. Please wait.

Presentation is loading. Please wait.

Lecture 12 Psyc 300A. Review: Inferential Statistics We test our sample recognizing that differences we observe may be simply due to chance. Significance.

Similar presentations


Presentation on theme: "Lecture 12 Psyc 300A. Review: Inferential Statistics We test our sample recognizing that differences we observe may be simply due to chance. Significance."— Presentation transcript:

1 Lecture 12 Psyc 300A

2 Review: Inferential Statistics We test our sample recognizing that differences we observe may be simply due to chance. Significance level (alpha) is the risk we are willing to assume that we will say there is a relationship between variables when in fact there isn’t (it’s due to chance).

3 Review: Type I and Type II Errors Accept the Null Hypothesis Reject the Null Hypothesis Null is really True Correct Decision Type I Error Null is really False Type II Error Correct Decision

4 Review: t-test SPSS Output

5 Experimenter and Participant Effects Participant Effects –Participants are not passive –Participant reactivity: behavior changes when participants know they are watched (may respond with cooperation, antagonism, social desirability) –Respond to demand characteristics

6 Social Desirability The pressure that participants feel to respond as they think they should, not as they actually feel or believe. The acceptable or PC response

7 Demand Characteristics These are cues that come from the experimenter or the experimental situation that tell a participant about the purpose of the experiment Participants may be right or wrong in their guesses Participants may change their behavior (e.g., become more or less cooperative)

8 Experimenter Effects Experimenter bias occurs when the behavior of the experimenter in some way affects the results of the study –Experimenter expectancy effects –Experimenter attributes

9 Experimenter Expectancy Effects When experimenter’s knowledge or belief about participants cause participants to act different from normal May involve –Demand characteristics (leaking) –Interpretation of responses –Subtle differences in behavior toward participants

10 Experimenter Attributes When characteristics of experimenter affect participants Examples: Age, ethnicity, attractiveness, gender, extraversion, controlling May limit external validity

11 Controlling Participant and Experimenter Effects Deception Blind studies Automation

12 Threats to External Validity Generalizing to populations Generalizing from lab settings Review: Replication –Literal (exact) –Conceptual

13 Review: Types of Experiments Between-Subjects (or Between- Participants) Design –Different subjects are assigned to each level of the IV –Random assignment to conditions –Difference between random assignment and random sampling Within-Subjects (or Within-Participants, or Repeated Measures) Design –Same subjects in all levels of the IV

14 Within-Subjects Designs Advantages –Participants serve as own controls Groups are equivalent at beginning Reduced variability among participants –Need fewer participants –More powerful (more likely to detect relationships when present) Limitations –Order effects –Can’t use when participant variables are IV –Sometimes long time lags (esp. longitudinal designs)

15 Order Effects and Counterbalancing Order Effects –Practice effects –Fatigue effects –Carryover effects Counterbalancing –Example: A-B-C –Complete –Partial

16 Matched-Participants Design Match participants on an important variable Different participants in different conditions (like between), but analyzed like a within- subjects Advantages: reduce demand characteristics, eliminate order effects, reduce variability Disadvantages: Finding matches, only matches for some characteristics

17 Developmental Designs Cross-sectional design –Between subjects design –Cohort effect Longitudinal design –Within subjects design Sequential design –Mix of cross-sectional and longitudinal


Download ppt "Lecture 12 Psyc 300A. Review: Inferential Statistics We test our sample recognizing that differences we observe may be simply due to chance. Significance."

Similar presentations


Ads by Google