1/53
Looks like no tags are added yet.
Between subject
each participant is exposed to 1 level of IV
Within subject
each participant is exposed to all levels of IV
Random assignment
assigning participants to certain conditions is random
Posttest-only design
Randomly assign ppl to two different conditions
Manipulate IV and measure DV posttest
Pretest-posttest design
Randomly assign ppl to different conditions
Measure DV (And other variables) pretest
Manipulate IV and measure DV posttest
Advantages of pretest-posttest design
Gives idea of what participants are like before IV
Can be used to check if random assignment yields equivalent groups
Even if it doesn't, there are statistical methods to account for pretest performance
Can be used to select ppl for research on special populations
Allows examination of “change” from pretest to posttest
Disadvantages of pretest-posttest design
Requires more time and effort
Participants may figure out what the experiment is about from the pretest, which can lead to ungenuine responses (solution=disguising pretest)
Matched pairs design
study participants are matched based on key variables/shared characteristics relevant to the topic of the study
Counterbalancing
systematically varying the order of measures/conditions in a study
You can counterbalance two conditions to make multiple conditions
3 things to do with time interval between treatments
Filler tasks can reduce learning/practice effects
Rest period can reduce fatigue and reduce the lasting effects of early treatment
Long intervals between tasks could increase dropout rate
Simultaneous within-subject design
used when participants make multiple responses in each condition and different IV levels are not blocked
Factorial Designs
designs with more than one IV (or factor) where all levels of each IV are combined with all levels of the other IV (s)
Main effect
the effect each variable has by itself, irrespective of the effect of the other variable
Ex: does caffeine affect exam performance overall?
Interaction effect
the effect of one IV depends on the particular level of the other IV
Simple effects
the effect of one IV on the DV at particular level of other IV (only relevant when interaction effects are present and we have p values for each condition)
Does caffeine affect exam performance among students who slept 4 hours?
Mixed factorial design
In the middle of between-subject and within-subject factorial design
Ex: students are assigned to either caffeine or no caffeine, but everyone takes a test after both 4 and 8 hours of sleep
Quasi-experimental design
study the effects of IV(s) on DV when control features of true experimental designs cannot be achieved
Either a control group is missing or participants aren’t randomly assigned
Helps to establish temporal precedence because the intervention typically occurs before the DV is measured
Often conducted to evaluate the effectiveness of a treatment
Quasi-experimental design 5 threats to internal validity
History: any event occurring between the pretest and posttest that is not the manipulation
Maturation: participants change over time
Testing: taking the pretest changes people's behavior
Instrumental decay: measuring instruments change over time
Regression toward the mean: scores tend to become closer to the average levels of the population when tested again
May occur when participants were selected because of extremely high/low scores on some variables
Nonequivalent control group designs
when random assignment is not possible
Groups already exist, groups may differ before intervention (confound potential)
Single Subject Experiment Design
quantitative research that studies in detail the behavior of a small number of ppl (1-10)
Determine whether a treatment or manipulation works on a single research participant
Individual behavior, clinical interventions, or educational strategies
Ex: reducing classroom interruptions….. Baseline= recording how often a child interrupts… intervention= introduce a reward for raising hands)... measurement= to see whether interruptions decrease
AB Design
baseline DV measurement, then treatment DV measurement
ABA Design (reversal design)
baseline DV, treatment DV, remove treatment DV
ABCA Design (multiple treatment reversal design)
baseline, treatment 1, treatment 2, baseline (withdrawal)
ABAB Design
baseline, treatment, baseline, treatment
Multiple baseline design
measure DV consistently but introduce treatment/manipulation to the participants at different time points, different skills, or in different settings
3 different variations for multiple baseline design
Varies across people (Diff times), across behaviors, across settings
Visual Inspection
examine graphs for changes in behavior across baseline and treatment phases
Statistical analyses
compare repeated measurements within the same person across phases, rather than comparing groups of ppl
Developmental research
study of human growth and development as a function of age
Ethical concerns about developmental research
respect each child as an individual in his/her own right
Respect relationship between the child and their family
Facilitate growth and development of each child
Use information appropriately, respect privacy, and maintain confidentiality
Cross-sectional method
individuals of different ages are studied at one-time point
Issues with cross-sectional method
Cohort effects: variations in characteristics among cohorts of ppl due to shared experience such as time of birth or years of a specific exposure (Ex: baby boomers and technology)
Longitudinal Method
same group of people is observed at different points in time as they grow older
5 Issues with Longitudinal Method
Attrition: people discontinue their involvement in the middle of the study
Resources: expensive and difficult
Limited to one cohort
TIme in history effects confounded with age changes
Possibility of practice effects
Sequential method
begins with cross-sectional method and follow cohorts longitudinally
Qualitative research
generates textual or visual data, primarily through the use of participant observation, in depth interviews or focus groups… Emphasizes context and meaning behind actions
Interviews
(structured or semi)
Structured used for surveys or studies that aim for a high level of standardization and comparability
Asked the same set of questions in the same order
Focus groups
(5-7 ppl; generally similar but dont know one another)
Strengths for qualitative research
rich detailed data, flexibility, contextual understanding
Challenges for qualitative research
subjectivity (potential for researcher bias), time-consuming, generalizability (may be hard for small populations)
Thematic analysis
identifying patterns and themes in data (most common data analysis)
Familiarization (immerse yourself in data by reading)
Generating initial codes
Searching for themes
Reviewing themes
Defining/naming themes
Writing up (integrate themes into a narrative)
Content analysis
systematic coding and categorization
Narrative analysis
examine stories and personal accounts
Grounded theory
developing theories based on data
Purposive Sampling
researchers select ppl based on specific criteria that are relevant to the study’s objectives
Snowball sampling
participants refer others that are similar to themselves on x dimension… assume participants have social ties along the dimensions of interest… Good for accessing “hidden” populations
Convenience sampling
select whomever/whatever you can find… should be sampling method of last choice
Participatory Action Research (PAR)
involves researchers and participants working together to understand a problematic situation and change it for the better
5 Levels of community involvement for PAR
informed, consultation, involvement, collaboration, empowerment
Goals of PAR
Empowerment, social change, collaboration
Principles of PAR
inclusivity, reflective practice, iterative process
5 Methods of PAR
photovoice, surveys, interviews, focus groups, journal entries
Strengths of PAR
community empowerment, relevance, sustainability
Challenges of PAR
power dynamics (between researcher and participant), time intensive, potential for conflict (different agendas among ppl)