EXP PSYCH L2 BASICS OF EXPERIMENTS

true experiment

has 4 elements:

1. manipulation

2. control

3. random selection

4. random assignment

manipulation

element of environment or the environment itself is purposely changed by experimenter

control

prevent extraneous / nuisance factors from contaminating experimental outcome

random selection

• getting small sample from huge population

• everyone should have an equal chance of being part of the sample

• if population is too small, it becomes the sample itself

• ex: fishbowl, roulette

random assignment

• happens after participants are selected

• participants are assigned in a random way

• ex: coin flip

quasi experiments

• esembling / resembles experiments but lacks elements to meet criteria for true experiments. thus, not a true experiment

- may lack control due to non random assignment & non manipulated independent variable

- may not include a control group

true vs quasi experimental design

true:

• cause & effect relationship

• offer more control = accurate outcomes / claims

quasi:

• lacks control needed to demonstrate cause & effect relationship

types of quasi experiment

1. one group designs:

• post test only

• pre test and post test

2. nonequivalent control group design:

• post test only

• pre test and post test

3. time series designs:

• basic

• interrupted

• control

one group post test only

• one shot case study / weakest design = lots of flaw

• experimenter measures dependent variable for just one group of participants after intervention / treatment

• weakness: can't tell if the intervention had effect bcs there's no comparison to how things were before it happened. It only looks at the results afterward.

one group pre-test & post test

• only one group

• measures dependent variable before & after intervention = can compare if intervention has effect

• weakness: could cause threats bcs there's no "control group" to compare with

control group

set of participants who does not receive treatment / intervention being studied

nonequivalent control group post test only

• researcher measures dependent variable (results) after treatment in one group & compares it to a nonequivalent control group.

• quasi-independent variable: participants are grouped based on their existing characteristics.

• weakness: any difference seen in results between groups might be bcs the groups were different to begin with = hard to know if treatment cause change / effect

nonequivalent control group pre test and post test

• compares pre test & post test results of 2 groups (ppl who received and didn't received intervention) = track changes over time

• weakness: no random assignments for participants = pre-existing difference could affect outcome

basic time series design

• uses series of observations over time, before & after intervention in one group = researcher can see pattern of changes

• using time in the study to eliminate problem associated with snapshot of behavior (looking at only one moment)

• weakness: won't work well for dependent variables that don't easily change (ex: IQ / personality)

interrupted time series design

• measures a dependent variable multiple times before & after a naturally occurring event / intervention (no manipulation)

• one group = no control group to compare

• goal: focuses on the changes of the same group over time

• weakness: limited control / no manipulation

control series design

• measures dependent variable multiple times before and after naturally occurring event / intervention (no manipulation)

• two non equivalent groups & no random assignment (ppl who received & didn't received intervention)

• goal: to compare both groups

• weakness: limited control / no manipulation

summary of types of quasi experimental design

types of true experiments

1. pre test and post test control group design

2. post test only control group design

3. solomon four group design

pre test and post test control group design

• measures pre test and post test of 2 groups, randomly selected & assigned participants for each group

• experimental group = receive tested intervention

• control group = different intervention or no intervention given

post test only control group design

• giving a randomly assigned experimental & control group a post-test after an intervention.

• post-test only helps eliminate test effects (recalling / memorizing info) of having both a pre-test and post-test.

• basically, post test prevents participants from recalling info from pre-test bcs it could influence their performance in pre test.

• helps to focus on just the results after intervention, without worrying about any effects caused by the timing of testing.

solomon four group design

• complex experimental design w/ randomly assigned 2 control and 2 experimental groups

• first pair of group: receives both pretest and post test

• second pair of groups: receives only post test

• in simple terms, SFGD combines the 2 previous design

• helps address better timing and prevent pre test memories from influencing results

SFGD example