L3 - Developmental Research Methods
→ Causality and developmental research (problem in psychology)
interested in cause and affect, want to make these causal claims (which most times we cant do)
There may be more causes than just the one we are interested in
we dont know the relationships between the potential causal factors
We usually find links that are positively correlated to make causal claims
ways in which we draw causal connections between 2 variables
Important to control correlated variables
typically the goal of an Experiment
Example of genetics affecting abilities
We want to be able to understand the causal connections between early experiences and later behavioural functioning
But CONFOUNDING variables (confounders) limit the extent to which we can truly assess cause and effect
Errors in making causal inferences
Selection bias - limitation in selection in London
Simultaneity bias - time period
Factors to consider in causal inferences
Randomization
Randomly sample to make causal inferences to avoid “convenient samples”
internal vs. external validity - IQ example
Does it measure what it claims to measure?
Causal inferences and growth
How can we account for growth and change in making causal inferences?
Problems of validity are problematic for growth and change perspectives.
We want to know what is changing and what is generating that change - we want to make a causal connection.
Contexts of Developmental Research
- What are methods of collecting data?
Descriptive Research
Research based on recording things that are observed
Often not causal (cant control something in the environment)
Not trying to manipulate anything or assess anything
→ E.g. Self report measures - Interviews and questionnaires, Teacher reports
Positives: lots of data, really fast
Negatives: Bias (how questions are formulated, can affect the answer)
Other types of descriptive research: → Observational Measures
Naturalistic Observations (Just observing people un an uncontrolled setting)
Positives: Good external validity
Negatives: Control, limited range of behaviours
Structured Observations (Scripted set of things happening, scenes that people come into)
Positives: control
Negatives: less descriptive, less external validity
Case studies
Detailed descriptive study of a single individual
Positives: Allows study for unusual situations, raises questions for further study
Negatives: Lack of generalizability
Feral children example
Non-descriptive approaches
Experimenter exerts some level of control to the study
Tries to measure different factors in different dimensions
→ Quantitative
Using numerical information
Countable
Measureable
“How many, “how much”, How often”
Often used in nomothetic approaches
→ Qualitative
Descriptive to some degree
Interpretations
Descriptive vs relational
Why, How, What happened
Often used in ideographic approaches
Correlational Research Approaches
Examines the relationship between two or more variables
Positives: Quantify relationships and make predictions
Negatives: no causality
Experimental Research
Manipulation of IV to examine effect on DV with random assignment
Positives: Causality
Negatives: not always possible to control all confounding variables; generalization becomes a problem with too much control
Quasi-experiments
Comparison of groups that differ on a characteristic of interest
e.g. performance of 2 classes that are enrolled in the same courses
Schools tested using generalized testing
no random assignment
Positives: Allows for comparison of variables that cannot be controlled
Negatives: No causality
→ Longitudinal designs
Subjects studied at different times to see of anything changed
E.g.
Studied at 2 years old
Studied at 4 years old
studies at 6 years old
studied at 8 years old
→ Cross sectional design
All assessments done in one moment in time with all age groups to see if they differ on a particular dimension
E.g.
Subject A - 2 years old
Subject B - 4 years old
Subject C - 6 years old
Subject D - 8 years old
→ Longitudinal-Sequential Design (aka Cross-sequential design)
Combines cross-sectional and longitudinal research
Children of different age groups are followed longitudinally
Hodges et al. (2021)
→ Causality and developmental research (problem in psychology)
interested in cause and affect, want to make these causal claims (which most times we cant do)
There may be more causes than just the one we are interested in
we dont know the relationships between the potential causal factors
We usually find links that are positively correlated to make causal claims
ways in which we draw causal connections between 2 variables
Important to control correlated variables
typically the goal of an Experiment
Example of genetics affecting abilities
We want to be able to understand the causal connections between early experiences and later behavioural functioning
But CONFOUNDING variables (confounders) limit the extent to which we can truly assess cause and effect
Errors in making causal inferences
Selection bias - limitation in selection in London
Simultaneity bias - time period
Factors to consider in causal inferences
Randomization
Randomly sample to make causal inferences to avoid “convenient samples”
internal vs. external validity - IQ example
Does it measure what it claims to measure?
Causal inferences and growth
How can we account for growth and change in making causal inferences?
Problems of validity are problematic for growth and change perspectives.
We want to know what is changing and what is generating that change - we want to make a causal connection.
Contexts of Developmental Research
- What are methods of collecting data?
Descriptive Research
Research based on recording things that are observed
Often not causal (cant control something in the environment)
Not trying to manipulate anything or assess anything
→ E.g. Self report measures - Interviews and questionnaires, Teacher reports
Positives: lots of data, really fast
Negatives: Bias (how questions are formulated, can affect the answer)
Other types of descriptive research: → Observational Measures
Naturalistic Observations (Just observing people un an uncontrolled setting)
Positives: Good external validity
Negatives: Control, limited range of behaviours
Structured Observations (Scripted set of things happening, scenes that people come into)
Positives: control
Negatives: less descriptive, less external validity
Case studies
Detailed descriptive study of a single individual
Positives: Allows study for unusual situations, raises questions for further study
Negatives: Lack of generalizability
Feral children example
Non-descriptive approaches
Experimenter exerts some level of control to the study
Tries to measure different factors in different dimensions
→ Quantitative
Using numerical information
Countable
Measureable
“How many, “how much”, How often”
Often used in nomothetic approaches
→ Qualitative
Descriptive to some degree
Interpretations
Descriptive vs relational
Why, How, What happened
Often used in ideographic approaches
Correlational Research Approaches
Examines the relationship between two or more variables
Positives: Quantify relationships and make predictions
Negatives: no causality
Experimental Research
Manipulation of IV to examine effect on DV with random assignment
Positives: Causality
Negatives: not always possible to control all confounding variables; generalization becomes a problem with too much control
Quasi-experiments
Comparison of groups that differ on a characteristic of interest
e.g. performance of 2 classes that are enrolled in the same courses
Schools tested using generalized testing
no random assignment
Positives: Allows for comparison of variables that cannot be controlled
Negatives: No causality
→ Longitudinal designs
Subjects studied at different times to see of anything changed
E.g.
Studied at 2 years old
Studied at 4 years old
studies at 6 years old
studied at 8 years old
→ Cross sectional design
All assessments done in one moment in time with all age groups to see if they differ on a particular dimension
E.g.
Subject A - 2 years old
Subject B - 4 years old
Subject C - 6 years old
Subject D - 8 years old
→ Longitudinal-Sequential Design (aka Cross-sequential design)
Combines cross-sectional and longitudinal research
Children of different age groups are followed longitudinally
Hodges et al. (2021)
