Please enable JavaScript.
Coggle requires JavaScript to display documents.
Preliminary Issues Regarding Research (cont) (Effect of people studying…
Preliminary Issues Regarding Research (cont)
Internal Validity
extent to which we can say that any effcets on the DV were caused by IV
Threats
sequence effect
- progressive/order (1st/last has effect); carryover effect (pairing)
selection
:
Non-equivalent
groups in diff conditions
confounds
: another variable that varies systematically with IV and could affect DV
history
: another event coincide with a study and affect its results (smth happened in btw the time of exp, micro/macro)
maturation
: change in participants as a result of time (practice effect, fatigue)
Regression towards the mean
: scores very diff frm mean at first measurement become nearer to mean in 2nd measurement (e.g. students who failed the course)
Experimental Design & Control
Terms
extrabeous variable - any variable (other than IV) that has potentail to influence the results
Random allocation
- differetn form random sampling (select sample from population)
Ex post facto
- x causal conclusion, cuz x manipulation → concluded as having different
Aim: inferring causation
good experimental design have good internal validity
randomly allocating, manipulating IV, treating all conditions equally except for IV, using appropriate control conditions
Experimental
- random allocation of (sufficient) Ps
Disad
- large smaple size, individual diff
Pre-experimental or quasi-experimetnal
- Ps are not randomly allocated to conditions or levels of the IV
Control
merely having repaeated measures in the control condition will control for the possible effcets of repeated testing such as testing and regression to the mean
think about potential alternative explanations for anticipated effects and control
Factorial designs
- 2/> categorical IVs are manipulated if it's a quasexp and all combinations of the IVs are tested
Within-subjects/repeated measures
Mixed design
Between-subjects
Additional control
Counterbalancing
- equal representation of the possible order of presentation of conditions in a repeated measures
with-groups- each participants go through 1 treatmetn sequences
within-subjects- present participants all treatment sequences
Blind design
- single/ double
Manipulation check
cover story
- deception for study's purpose
Placebo conditions/ non-specific treatment control groups
- no-treatment control/ waiting-list control (no treatemtn during exp, but, was givena fter exp)
Matched groups designs
- participants in two (or more) conditions are matched on potential, relevant extraneous variables (problem: inequavalent participants in brw-subjects)
How do we control
apply similarity in all conditions (random assign, appropriate control, treat the same except for IV)
Effect of people studying people
Experimenter bias
- inadvertently administer experimental instructions differrently in differetn conditions or when discrietion/judgment is used in determining scores (unintentiona/intentional- be more aware)
Attribution
- mortality, drop outs (not necessarily a problem depends on the pattern anf possible cause)
Non-specific treatment effect
- similar to placebo effects
Placebo effect
- participants' expectations about what will happen
Demand effect
- features of a study that inform the participants about the study's purpose, result in participants knowing/guessing what the researcher expects (good participants effect)
Testing effect
- changes that occur in a measure/score over 2+ testing occasions because of the earlier measurement (practice effect)
Observe reactivity
- chamges in participants' behavior owing to abservation/ measurement
Social desirability
- create good impression in relation to issues that have a siocial/moral value
Instrumentation
- tool/equipment is interpreted