CH. 2 Flashcards
Three Goals of Science
Describe
- Take a photograph of the state of events. How are things currently?
Means, variance, correlations, percentages, frequencies
- Predict
Based on past observed patterns, who do we think will do well in this job?
Take correlations (and regressions) from one sample in which we KNOW the answer and apply the numbers to a new sample in which we don’t
- Explain
Most difficult—why did Sally have the highest sales last year? What’s her secret?
Seeks a causal answer to what happened
Requires scientists to form a theory and test that theory in an experiment (to rule out other possible explanations).
Usually compare group with t-tests, F-tests, or regression
The Empirical Research Process
- Statement of the Problem
- Design of research study
- Measurement of variables
- Analysis of data
- Conclusion from research
Statement of the Problem: What is a Theory?
- A statement that proposes to explain relationships among phenomena
Example: Attractive people are more likely to get a job than unattractive people
Inductive vs. deductive method
Theory comes first in deductive method
Form a theory, test that theory
Data come first in inductive method
Gather data, form a conclusion based on that data
Why is Theory important and why might it not be necessary? Ups and Downs?
Why theory is important in psychology
- Integrates and summarizes past research and provides a framework for new research
Reduces the tendency toward Type I error (saying you found something that doesn’t actually exists)
Why theory may not be necessary (or at least not always)
- People can miss important findings if they are only looking where they’re looking (become blinded by their theories)
If we never did exploratory analyses, how would counterintuitive (unexpected) findings be found?
- A lot of effort is expended forming theories that may not be supported or will eventually fall out of favor
Productive research does not require a theory
Internal vs. external validity
Research design
- A plan for conducting a study
Choice of design depends on the nature of the problem being studied (and cost and feasibility)
Trade-off between internal and external validity
- The more controlled the experiment is (the more you control for potential third variables or confounds) the more internal validity you have
The more natural the setting, the more external validity you have
- The more likely it is that the results in your study are representative of what would happen in the real world (how generalizable is it to the real world)
Primary Research Methods (PRM)- 4 Parts
1. Laboratory Experiment
1. Laboratory Experiment People come into a highly controlled laboratory setting and get randomly assigned to condition Can infer causality (more internally valid) Less generalizable (less externally valid)
PRM - 2. Quasi-Experiment
Something about the study is not randomly assigned—maybe you’re comparing men and women or comparing different work groups
You could randomly assign a particular condition, but if another is not randomly assigned, it’s still a quasi-experiment
Book says all field studies quasi-experimental
Less control, more generalizability
PRM - 3.
No manipulation—everyone gets the same questions
Response rate problem
Not everyone responds, might decrease generalizability
Truthfulness of responses
For a job, people may want to lie to look good
However, can make surveys anonymous to increase truthfulness
Internet surveys
PRM - 4. Observation
Generate ideas for further testing
Rich with data in environment where behavior being studied occurs
BUT your presence could affect behavior
Secondary Research Methods (SRM) - 1. Meta-analysis - Problems with it.
Meta-analysis
Statistical study of studies – aggregates results
Estimate of “true” relationship
Subjective decisions have to be made
File drawer effect of non-published studies
Problem from differences in level of analysis of original studies
If one study focuses
SRM - 2. Data Mining
“Big Data”
Used to reveal patterns of large data sets
Qualitative data
Qualitative data can be collected within the context of any of the contexts mentioned (laboratory, quasi-experimental, questionnaire, observation) Book says qualitative research is a class of research methods in which the investigator takes an active role in interacting with the subjects he or she wishes to study. I think qualitative research is anything that is not automatically put into numbers (e.g
Ethnography
The art and science of describing a group or culture (e.g., work group or organization)
Researcher tries to keep an open mind but nearly impossible to not have preconceived notions
Tries to be aware of what is the participants’ awareness and experience versus the detached outside investigator
Emic – insider perspective
Etic – outsider perspective
Still rarely used in comparison to quantitative research
Steps 3&4 - Measurement & Analysis: Quantitative
Measurement of Variables
- Inherently numerical
Can be continuous (like age, time, speed, and weight)
A “ratio” scale is a continuous scale that has a zero point that means nothing is there (e.g., 0 lbs)
Can be interval
Likert scales (1 = not at all, 7 = Extremely)
