Scribd
Upload
5 views

Expe Midterms

Uploaded by

mh.taylor.alexa
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
5 views

Expe Midterms

Uploaded by

mh.taylor.alexa
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

• Non-Experimental Designs: Survey Research

• Key Concepts:
• Survey Research: Collects data via questionnaires/interviews to study opinions, attitudes, and
behaviors.
• Advantages:
o Efficient for large datasets.
o Anonymity improves honesty for sensitive topics.
o Complements experiments by identifying correlations.
• Limitation: Cannot establish causal relationships (no variable manipulation).

• Constructing Surveys
• Steps:
o 1. Define Objectives: Specific goals (e.g., "Measure psychology students’ attitudes
toward animal research").
o 2. Question Types:
▪ Closed (Structured): Fixed responses (e.g., yes/no, multiple-choice).
▪ Open (Unstructured): Free-text responses.
o 3. Data Analysis Plan:
▪ Closed: Quantitative (percentages, counts).
▪ Open: Qualitative (content analysis for themes).
• Question Design Tips:
o Avoid ambiguity, double-barreled questions, and non-exhaustive options.
o Use neutral wording (e.g., "terminate pregnancy" vs. "kill unborn babies").

•
• Measuring Responses
• Scales of Measurement:
o Nominal: Categories (e.g., gender, ethnicity).
o Ordinal: Ordered ranks (e.g., beginner/intermediate/fluent).
o Interval: Equal intervals, no true zero (e.g., IQ, temperature).
o Ratio: True zero (e.g., height, time).
• Preference: Use the highest possible scale (interval/ratio) for richer statistical analysis.

•
• Important Considerations for Survey Items
• First Question: Relevant, easy, closed.
• Avoid:
o Leading/value-laden questions.
o Context effects (order of questions influencing responses).
• Response Biases:
o Social desirability: Answering to appear favorable.
o Yea/Nay-saying: Agreeing/disagreeing regardless of content.
o Position preference: Choosing based on answer order.

•
• Collecting Survey Data
• Methods:
o Interviews:
▪ Structured: Uniform questions (quantitative focus).
▪ Unstructured: Flexible (qualitative focus).
• Population vs. Sample:
o Population: Entire group of interest.
o Sample: Subset used to infer population traits.
• Key: Ensure representativeness for generalizability.

•
• Sampling Methods
• Probability Sampling (Random Selection):
o Simple Random: Equal chance for all (e.g., lottery).
o Systematic: Select every nth individual.
o Stratified: Proportional subgroups (e.g., academic years).
o Cluster: Randomly select groups (e.g., cities).
• Nonprobability Sampling (Non-Random):
o Convenience: Easy access (e.g., psychology class).
o Quota: Match population proportions (non-random selection).
o Purposive: Specific subgroups (e.g., sales employees).
o Snowball: Referrals from initial participants.

• Alternatives to Experimentation: Correlational and Quasi-Experimental Designs


• Quasi-Experimental Designs
• Definition: Resemble experiments but lack manipulation of antecedent conditions and/or
random assignment. Inclusion in treatment groups is based on preexisting conditions.
• Purpose: Explore differences between groups or compare treatment effects on preexisting
groups. Cannot establish causation.
• Examples:
o Impact of a natural disaster on children.
o Emotional reactivity after a bombing (exposed vs. unexposed).
o Gender differences in sleeping patterns (cannot confirm gender as the cause).
• Types:
o Ex Post Facto Studies: Examines effects of already existing subject variables on a DV
without manipulation.
o Nonequivalent Groups Design: Compares treatment effects on preexisting groups.
Useful in natural settings.
o Longitudinal Design: Measures the same group over time.
o Cross-Sectional Studies: Compares different developmental stages at one time.
o Pretest/Posttest Design: Assesses whether an event changes existing behavior
levels.
• Correlational Designs
• Definition: Determines the relationship between two traits, behaviors, or events.
• Uses: Explore behaviors not well understood.
• Examples:
o Extraversion and conscientiousness have a positive effect on academic achievement.
o Lower levels of formal education correlate with a higher risk of developing Alzheimer’s.
• Properties:
o Linearity: Linear or curvilinear relationship.
o Sign: Positive or negative (+ or -).
o Magnitude: Strength of the correlation (-1 to +1, closer to 1 is stronger).
• Scatterplot: Graphic display of data points illustrating correlation properties.
• Outliers: Extreme scores that disturb data trends.
• Caveats: Correlation does not equal causation.
• Reasons for Not Proving Causation:
o Causal direction (A could cause B or vice versa).
o Bidirectional causation (A and B affect each other).
o Third variable problem (a third variable influences both).
• Formulating the Hypothesis
• Hypothesis: A statement about a predicted relationship between variables.
o Nonexperimental Hypothesis: Predicts relationships between variables but not
causally.
o Experimental Hypothesis: Predicts the effect of an IV on a DV.
• Characteristics of a Good Hypothesis:
o Synthetic Statement: Capable of being true or false.
o Testability: Can be assessed by manipulating an IV and measuring the DV.
o Parsimony: Simple, avoids unnecessary assumptions.
• Models for Formulating Hypotheses:
o Inductive Model: Reasoning from specific cases to general principles.
o Deductive Model: Reasoning from general principles to specific predictions.
o Combined Approach: Induction to develop general propositions, then deduction to
make specific predictions.
• Building on Prior Research: Reviewing existing studies to identify unanswered questions,
suggest new hypotheses, identify mediating variables, recognize problems, and avoid
unintentional duplication.
• Serendipity: Discovering something unexpectedly.
• Windfall Hypothesis: Unexpected finding that is favorable.
• Types of Hypotheses:
o Directional Hypothesis: Predicts the direction of the difference between groups.
o Nondirectional Hypothesis: Predicts that groups will have different values on the DV.

• Basics of Experimentation
• Key Concepts:
• Independent Variable (IV): Actively manipulated by the researcher (e.g., color of paper in a
test).
• Dependent Variable (DV): Measured outcome influenced by the IV (e.g., test scores).
• Levels of IV: Distinct conditions (e.g., blue vs. yellow paper).
• Confounding: Extraneous variables systematically affecting results (e.g., testing groups at
different times).
• Operational Definitions
• Experimental Operational Definition:
o Specifies how the IV is manipulated (e.g., "subjects fast for 12 hours" to define hunger).
• Measured Operational Definition:
o Specifies how the DV is measured (e.g., "reaction time in milliseconds").

• Evaluating Operational Definitions


• Reliability: Consistency of results.
o Interrater: Agreement between observers.
o Test-Retest: Consistency over time.
o Inter-Item: Consistency across survey items.
• Validity: Accuracy of measurement.
o Face: Does the measure appear valid?
o Content: Does it cover all aspects of the construct?
o Predictive: Can it forecast future outcomes?
o Concurrent: Correlates with existing measures.
o Construct: Measures the intended concept.
o Internal: Ensures IV causes DV changes.

• Threats to Internal Validity


• History: External events altering results.
• Maturation: Natural changes in participants (e.g., fatigue).
• Testing: Prior exposure affects performance.
• Instrumentation: Changes in measurement tools.
• Selection: Pre-existing group differences.

• Planning the Method Section


• Participants: Who was studied?
• Apparatus/Materials: Tools used (e.g., specialized equipment).
• Procedure: Step-by-step process for replication.
• Controlling Extraneous Variables
• Physical Variables:
o Elimination: Remove distractions (e.g., soundproof room).
o Constancy: Keep conditions identical (e.g., same time of day).
o Balancing: Distribute variables equally across groups.
• Social Variables:
o Demand Characteristics: Cues influencing behavior (controlled via single-
blind designs).
o Placebo Effect: Improved performance due to expectations.
o Experimenter Bias: Researcher’s behavior affecting outcomes (controlled via multiple
experimenters).
• Context Variables:
o Avoid biased samples (e.g., experiment titles attracting specific participants).

You might also like