chapter 14 Flashcards
(35 cards)
two epistemological approaches
Nomothetic Approach – Group Designs
**Idiographic Approach – Case Studies / Single-Subject Designs
Nomothetic Approach – Group Designs
Goal: discover general principles that apply across many people or cases
Used when the groups of cases are the unit of analysis
Example: a correlations study of 500 high school students to discover the screen time effects on anxiety using survey data
Descriptive – doesn’t provide a lot of information
**Idiographic Approach – Case Studies / Single-Subject Designs
Goal: understand a specific individual or case to obtain an in-depth data
Used when the individual case is the unit of analysis
Example: a single case study of a high school student to understand screen time effects on anxiety using an interview
Explains thing
Exceptions to Research Findings
Behavioral science is probabilistic
Research findings uncover generalities and trends
But, there are always exceptions!
Expectations do not invalidate research findings, but should they be ignored?
Criticisms of Nomothetic Approach (i.e., group designs)
- No effort to explain error variance
- Do not address the question of how many of the participants in a study were affected by the IV
- Do not try to see whether one can repeat the effect of the IV within a single study
Group vs Single-Case (1) ERROR VARIANCE –> group design argument
Averaging across participants provides a more accurate estimate of a variable’s general effect
Group designs allow us to estimate the amount of error variance in our data
In a study on screen time and anxiety, some participants show very high anxiety while others don’t. The average shows an effect, but researchers don’t investigate other factors that might explain why some individuals react differentially to screen time
Group vs Single-Case (1) ERROR VARIANCE –> single-case argument
Error variance is created by averaging over participants in a group design (inter-participant variance)
Researchers using group designs ignore the “real” error variance within the participant
This intra-participant variance may be more important to understand
A student shows increased anxiety after 3 hours of tik tok use. A single subject design might track their mood, screen content, time of fuse, and even sleep the night before to pinpoint the source of variation. That’s way more personalized and explanatory than just calling it “error”
Group vs Single-Case (2) GENERALIZABILITY –> group design argument
Averaging the scores of several participants reduces the idiosyncratic responses of any one participant to show the general effect
A study finds a statistically significant effect of a mindfulness app on stress, but doesn’t report that only 30% of participants actually improved – the rest had no change or got worse. The focus is on the overall average, not individual impact
Group vs Single-Case (2) GENERALIZABILITY –> single case argument
Averaging responses may not accurately describe any particular participant’s responses
Instead of saying “the app reduces anxiety on average,” a single-subject design shows that Student A’s anxiety score dropped consistently every time they used the mindfulness app – or didn’t. You know exactly who benefited and how reliably
Group vs Single-Case (3) RELIABILITY –>
Group design argument
Reliability of findings is established by replicating studies
We can show that using social media increases stress after a single session, but we don’t test if the same person shows increased stress every time they use it over multiple days. So we don’t know if the effect is reliable within subjects
Group vs Single-Case (3) RELIABILITY
single-case argument
Reliability of findings show be established via:
- Intra-participant replication
- Inter-participant replication
A teen is tracked for two weeks. During “A” phases (no social media), anxiety is low. During “D” phases (3 hours of social media), anxiety spikes. This on-off pattern repeats constantly – confirming the effect is not a fluke
Intra-participant replication
Replicating the effects of the independent variable within a single participant
Inter-participant replication
Seeing whether the effects obtained for one participant generalize to other participants in the same study
Single-Case Experimental Designs
Unit of analysis is not the experimental group, as it is in group designs but rather the individual participant
More than one participant may be studied, but their responses are analyzed individually
You study 3 teens using the same intervention (e.g., reducing screen time). You create 3 individual graphs to track each teen’s anxiety pattern across time. You do not average their scores or look for a “group mean.”
Data from Single-Participant Designs
Cannot analyze single case data with inferential statistics such as t-test and f-tests
Instead!!! Use:
Graphic analysis
graphic analysis
research usually inspects the graph of the individual participant to see if the independent variable had an effect
Criticized for having no explicit criteria for deciding when an effect significant
Single Case Designs
Single Case Experiments
- Classical/operant conditioning
- Psychophysiological process
- Behavior modification
- Demonstrational purposes
classical/operant conditioning
Pavlov’s dog
Skinner’s box: rat presses level for food reward
Psychophysiological process
Drug studies: measuring heart rate changes in a person after taking caffeine
Behavior modification
Techniques for changing problem behaviors based on operant conditioning
Demonstrational purposes
Ebbinghauss effect: a student memorizes nonsense syllables to show how memory fades over time
Case Studies and examples
- Source of insight and ideas
- Describe rare phenomena
- Psychobiography – applying concepts and theories from psychology in an effort to understand famous people
- Illustrative anecdotes
Examples:
Freud’s case studies
Jung’s case studies
Festingerr’s case study of cults
Basic Single-Case (list)
- ABA Designs
- Multiple-I Designs
- Multiple baseline designs
ABA Designs
ABA Design (reversal designs)
- Behavior is measured (Baseline period; A)
- Independent variable is introduced (B)
- Behavior is measured
- Independent variable is removed (A)
- Behavior is measured