Research Methods in Industrial Organizational Psychology
Introduction to Research Methods in Industrial Organizational Psychology
Overview
Chapter two focuses on research methods in Industrial Organizational (IO) Psychology.
Assumption of familiarity with statistical and research method terms.
Encouragement to ask questions if assumptions of prior knowledge are inaccurate.
The Scientific Method
Definition of the Scientific Method
A systematic procedure used to investigate phenomena, acquire new knowledge, or correct and integrate previous knowledge.
Relies on empiricism, a philosophical view that knowledge comes from sensory experience.
Stages of the Scientific Method
Different interpretations exist (4-9 steps).
Textbook Reference (5 Stages):
State the problem.
Design a research study.
Measure variables.
Analyze data.
Draw conclusions, influencing subsequent studies.
Characterized as a circular system.
Approaches to the Scientific Method
Inductive Approach
Collecting evidence first, leading to theory formation based on data analysis.
Common in applied IO psychology as practitioners often identify issues from real-world observations and seek explanations.
Deductive Approach
Starts with existing theories or assumptions, tests them through data collection.
More common in scientific, academic research.
Comparison of Approaches
Both are valid for knowledge development but differ in execution.
Designing Research Studies
Importance of Research Design
Prevents threats to internal validity and confounds, aiding in accurate measurement of variables.
Requires careful planning.
Key Stages in Planning (Inductive or Deductive):
Selecting methodology.
Defining measurement procedures.
Considering data analysis techniques influencing data collection methods.
Methodological Considerations
Naturalness vs. Control in Research Settings
Naturalness: Enhances external validity and generalizability of findings.
Control: Enhances internal validity, allowing for manipulation and ruling out confounding variables.
High naturalness -> Low control and vice versa; balance is crucial based on research questions.
Primary Methods of Inquiry
Experiments
High control, low naturalness.
Conducted in lab settings, allowing causal conclusions due to controlled variables.
Limitations in generalizability to real-world settings.
Quasi-experiments
Medium control, moderate naturalness.
Lack of true random assignment; useful for exploring causal relationships but risks internal validity issues.
Common in field research settings.
Questionnaires
Low control, moderate to high naturalness.
Measures self-reported information; critical for gathering large samples but subject to biases.
Reliabilities and validities must be assessed rigorously for accurate results.
Direct Observation
Low control, high naturalness.
Purely descriptive; observing without intervening, poses issues like observer effects.
Suitable for small sample sizes, often used in job analyses.
Secondary Methods of Inquiry
Meta-analysis
Statistical technique combining results from multiple studies to discern patterns and overall effects.
Offers more objectivity than individual studies, balancing researcher biases.
Challenges include range restriction and the file drawer problem, as non-significant studies may remain unpublished.
Data Mining (Big Data)
Relational method analysis of extensive data sets to uncover relationships and correlations.
Notable for generating insights into organizational trends.
Example: Browser preference tied to applicant quality; indicates conscientiousness.
Other Methods of Inquiry
Organizational Neuroscience
Combines neuroimaging and workplace behavior studies, still emerging in IO psychology due to high costs.
Potential for growth as technology becomes more accessible.
Qualitative Research Methods
Focus on uncovering richer information rather than numerical categorization.
Often used for cultural analysis in organizations but typically involves small sample sizes.
Conceptualizing Variables in IO Psychology
Definition of a Variable
A non-constant measure that can take on different values. Ex: sex, trained/untrained.
Types of Variables:
Quantitative Variables: Numeric values reflecting measured constructs (e.g., performance scores).
Qualitative Variables: Categorical distinctions (e.g., trained vs. untrained).
Key Variable Classifications
Independent Variables (IV)
Manipulated by the researcher in experiments/quasi-experiments.
Cannot include preexisting characteristics of subjects.
Dependent Variables (DV)
Variables expected to be influenced by the IV, typically performance or attitudes in organizational settings.
Predictor vs. Criterion Variables
Predictor Variables: Used to predict outcomes (similar to IV).
Criterion Variables: Outcomes of interest; can be similar to DVs but may include broader variables in predictive models.
Conclusion and Further Reading
Encouragement to review textbook for statistical methods: correlation, analysis of variance, descriptive statistics, regression techniques.
Invitation to clarify any questions during office hours or via email before progressing to the next chapter.