Ch. 3 Research Methods Flashcards
variable
a characteristic that varies from person to person
dependent variable
expected to change based on independent; an outcome affected by independent variable(s)
independent variable
manipulated in study
experimental design
control vs. treatment groups
ex. presence vs. absence of natural light in classrooms
- relatively confident in the inference made
quasi-experimental design
researchers compare groups on predetermined characteristics
- cannot use random assignment
ex. age, gender, sex, social class
single factor design
studies that catalog information about how people perform based on their age but do not attempt to rule out social or historical factor
- only one variable studied
age
number of years a person is on the Earth; chronologic measure
cohort
year of birth
ex. technology- Cohort effect
- normative graded influences
time of measurement
year of testing
ex. Spring 2023
- presently affecting people
ex. the economy affecting financial security, Spring 2020
- historically graded affects
longitudinal design
follows one group of people over time; cohort followed over time
- including at least two time points
- does not tell age differences
- age within group variability
-everyone is averaged into one score
- does not address time of measurement and cohort effects
attrition
→ mortality; reason for participant dropping out
- were they different in a fundamental way?
- a design limitation
longitudinal limitation
does not take into account for social or historical factors; inability to differentiate changes due to age and the effect of social or historical events
selective attrition
people are dropping out of different groups at different times
- averages who have died are no longer there making averages increases
- average becomes the score of the survivors
practice effects
scores become better due to practicing
cross-sectional design
conducted at a single point in time
- a score for 1960, 1950, 1940 cohorts
- outcome declines with each decade
- don’t know if it is a function of age, time of measure, cohort effect
- most frequent design used → relatively cheap and simple
- take into account environmental factors
sequential design
look systematically at age, time of measurement, cohort
- a sequence of studies
cohort sequential design
cohorts are compared at different ages
- age and cohort
time sequential design
the data are organized by age and time of measurement
- age effects at different times of measurement
cross sequential design
cohorts are examined at different times of measurement
correlational design
relationships are observed among variables as they exist in the world
- cannot determined cause and effect
correlation coefficient
indicates direction and strength of relationship
- closer value is to 0 is weaker; closer value is to 1 is stronger
- valence- positive or negative
correlation does not
equal causation
multivariate
more than IV, one DV
- allow us to identify a set of predictors that have a statistically significance to the outcome
- how variables work together
multiple regression
appropriate when there are many factors for one outcome
logistic regression
allows us to access the likelihood of an individual receiving a yes or no response
- correlational to categorical variables
mediation
refers to a third variable explaining the relationship between the independent and dependent variables
- cognitive process
- why is something happening?
moderation
when the effect of the variable depends on the effect of another variable
- under what circumstances does another variable affect another
- the affective age may depend on sex
path analysis
- arrows are us telling the program we expect what relationships
- a specific path wanted to be tested
- data either fits the model or does not
- similar to multiple regression by accessing multiple variables at one time
- difference is it can address relationships between independent variables
structural equation modeling
- latent variables- is not independent but comes from responses of other variables
- can be used with path analysis interchangeably most of the time
- only has to be done once
hierarchical linear modeling
- allows for modeling individual change over time
- allows us to look at many factors at once
- larger variables that impact us in the moment
qualitative studies
- providing a written response; responding to open-ended questions
- thinking, saying, and how they are interacting
- thoughts, feeling, emotions in the language they produce
archival research
- gathering data from medical records; it already exists
- cannot control quality of data
-records could be partially complete, damaged, not culturally appropriate questions
surveys
- allow us to access a lot of people relatively cheaply
- quantitative and qualitative factors
- can vary, flexible
- reach a large amount of people
- people might not be paying attention, could be confused, answer randomly, lying
epidemiological studies
- prevalence of diseases in populations or location
- public health statistics
- give connection to origin; rate of transmission
- prevalence and incidence
prevalence
estimate of percentage of who ever had symptoms in a given period
incidence
estimate of percentage who first develop symptoms in a given period
-first time a symptom had arisen
- long covid
case reports
- extremely detailed, look at rare cases
- oliver sacks: beautiful detail of individual cases
- not really valuable for greater population; will not apply to general population
- give new information about specific cases
focus groups
- qualitative methods, answer questions from facilitator
- get people to start talking about particular phenomenon
- useful when we do not have a good understanding of what we are dealing with
daily diaries
- used to use beepers
- when notification was received, had to write down a response
- allows us to model what individuals do in a specific day
- access multiple people at multiple times of measurement
- randomly assign data collection times
- capture intraindividual difference
observational
- what they look like in the real world
- structured, naturalistic, laboratory
- possibly to randomly select observations
- see patterns that would not be seen by another method
meta-analysis
- units are not from individual people but from individual studies
- sample = how many separate studies are reviewed
- get beyond about what one study suggested but looking at numerous
- confidence intervals give margin of error for observed effect
- studies assess the relationship between independent and dependent variables
- overlap of whiskers share converging findings in other studies
reliability
yields consistent results every time it is used
- expect similar outcomes
test-retest reliability
determined by giving the test on two occasions to assess whether respondents receive similar scores across both administrations
internal consistency
indicates whether respondents answer similarly on comparable items.
validity
the test measures what it is supposed to measure.
content validity
an indication of whether a test designed to assess factual material accurately measures that material
ex. breakfast & IQ
criterion validity
indicates whether a test score accurately predicts performance on an indicator measure, as would be used in a test of vocational ability that claims to predict success on the job
construct validity
used to assess the extent to which a measure intended to assess a psychological construct is able to do so.
ethics in research
- informed consent
- right to withdraw
- confidentiality
- debriefing
any form of deception - suggest resources
