Experimental Research Flashcards
(19 cards)
What are the two main experimental designs. (ATTENTION WE SAID EXPERIMENTAL NOT RESEARCH DESIGNS)
Between groups (experimental and control groups, look at difference between groups) and within groups (one group, every participant undergoes every treatment - assess differences in each participant).
Can we talk about the use of variables such as dependent and independent with correlational studies?
No.
What variable has levels? Give examples of levels.
The independent variable (the one being manipulated, and which we suspect affects the outcome (DV))
ex:
Does metal music affect levels of aggression differently in people who listen to it recreationally compared to those who do not?
IV: music type
two levels: jazz music and metal music (two treatment conditions that you wanna assess!!!!!) - IVm
IVp : relates to participants differences ( genre those who listen to metal music recreationally and those who dont, you dont manipulate this, you just take it into account cause might influence results)
DV: aggression levels
Ex: does cultural-specific knowledge affect other perception (clue affect=experiement !!!!)
IV: cultural background
Two levels: French or French-Canadian
DV: ratings of odors
Whats the idea of IVms?
You manipulate one variable to create two treatment conditions (two levels more la mais mtn 2), then you measure your dependant variable to obtain set of scores (genre aggression level scores). then you compare scores in treatment A with the scores in treatment B. same idea que ca soit ace within group or between group design.
A within-group design: Each participant experiences both conditions, and their scores are compared across the treatments.
A between-group design: Different groups of participants are assigned to each condition, and the scores of the groups are compared.
What are the four major elements of a true experiment?
Manipulation, measurement, comparison and control. As soon as you see an exp. research, know that all four major elements are present.
What does the element “manipulation” (1) refer to?
It consists of identifying the specific values (Levels) of the independent variable to be examined and then creating a set of treatment conditions corresponding to the set of identified values. This step allows us to determine the direction of a relationship (cause we are manipulating one variable) and then see the effect it has on the other (DV)
For exemple there is a relationship between ice-cream consumption and temperature (correlational design). However because of the directionality problem, we dont know which causes what, all we know is that there is a relationship. However, with a true experiment, we can manipulate one variable (lets say we manipulate room temperature ) and see the effect is has on ice cream sales. We can then conclude that yes changing in temperatures are responsible for causing changes in ice-cream consumption and not the other way around.
What does the element “Measurement”(2) refer to?
Refers to your DV. Construct which you op def. : tells you how you are gonna assess/measure that construct /variable. You then take measurements for each of your treatment conditions (IV levels!!!!)
What does the element “comparison”(3) refer to?
Well essentially it means you compare the scores measured in each of your treatment conditions. Whether it be with a between group design or within group design.
What does the element “control” (4) refer to?
Basically, it consists thinking ahead of time about variables that may influence your outcome (extraneous variables that could be confounding!) and control them to either eliminate or reduce their impact!
example:
You re conduction an experiment to see whether preschool children prefer sweetened or unsweetened cereal.
well if one of those cereal is colourful and the other not well then what are you really assessing? their preferences of taste or of color!!!!!
Remember ideally you want everything to be the same in your experiment (your subjects, the instruction, the environment conditions…) Bref EVERYTHING EXPECT THE TREATMENT CONDITIONS!!!!!!!!!!!!!!
What are extraneous variables?
An extraneous variable is any variable that you’re not investigating that can potentially affect the dependent variable of your research study. Becomes a confounding variable when affects the dependent variable(outcome), but is also related to the independent variable.
Hypothesis: If people are given a meal on a larger plate, they will eat more than if they are given a meal on a smaller plate. What is the IV and the levels (values of the IV) and the DV? What are som possible extraneous variables that you might wanna take into account?
IV: size of the plate
Levels: small and large plate
DV: consumption
extraneous variables:
hunger levels; have you eaten before this?
appetite
dietary preferences
if working with a within-group design: order effect ( if you start with big plate, what are the odds they are still hunger for smaller plate )
bref the idea with extraneous variable is that you wanna try and equalize as much as possible, do the only difference really is just the treatment received.
Why are literature searches nice?
They allow us to see what has been done before, what were the extraneous variables that were OR were not accounted for, so that you can adapt your experiment and make it better.
What does “experimental control” refer to?
Experimental control refers to the strategies and techniques researchers use to minimize the influence of extraneous variables in an experiment, ensuring that any observed changes in the dependent variable are due solely to the manipulation of the independent variable.
What are 3 categories of extraneous variables?
Environmental variables (temperature, room…), participant variables (ex; dietary preferences, age, gender, cultural background, Socioeconomic status, personality traits, cognitive abilities…) and time-related variables (Fatigue
Practice Effects
Order Effects
Carryover Effects
Maturation
History Effects
time of day…)
Are individual difference bound to happen?
YES!!!! there is always gonna be some variability.
for example the example with food big plate, small plate: you could have a big eater, a light eater, one who is on a diet, another who is hungry……….. you have to take this into account in order to [ for between group design] minimize within-group differences (non-systematic variance) and maximize between group difference (systematic variance) and for within-group design:
In a within-group design, individual differences have LESS impact compared to a between-group design because each participant serves as their own control.
Bref with extraneous variables, verify first if their impact could actually affect outcome, if yes then assess that!!!
What are 4 techniques to control extraneous variables?
- Removing confounds:
how could that work?
Ok with the example with pasta and plates, lets say you think having an eating disorder is a confounding variable to the study. What do you do? Well you FILTER YOUR PARTICIPANTS. So for instance, you can send them a questionnaire on whether they have an eating disorder or not, to make sure or you could also schedule a meeting face to face with them if you wanna be the one assessing whether that person has an eating disorder (psychologist or medical expert…)!!! cause that person might not be aware on their own….
OTHER EXAMPLE:
Lets say you think sex is a confounding variable. well then you are gonna have a harder time removing this confounding variable (not always faisable), you re gonna have to have a reallyyyyy good explanation.
- Holding variables constant
Some extraneous variables such as time of day (time-related extraneous variable) can be held constant all your treatment groups (con. not like you can remove it lol.
- Matching values ( matched-pair design)
A man in each gp, a woman in each
a stage 4 cancer in each group….
bref tu cree des paires first (man-man; woman-woman and then randomly assign with goes in which)
comes with randomization too!
- Control by randomization
Randomly assign individuals. Par default you always want randomization!!!!!
ex: randomized block design (blocks: ex high iq medium iq low iq, in each treatment conditions there will be those blocks and indivula should have been randomly assigned to one or the other)
or again just simple randomization
What are the two treatment condition groups you could have
Experimental group: treatment condition
Control group: no- treatment condition
- control group could also inclue placebo control condition!!!
example:
Does listening to Mozart make you smarter?
IV: type of music listened to
2 levels: not listening to Mozart or listening to Mozart
DV: IQ score? performance on a math test? bref whichever op def you choose!
experimental: listens to mozart
control: no music (or could be Justin Bieber or whatever musician…)
example:
Does new medication help relieve depressive symptoms/
with this one you conduct a clinical trial, you can have a placebo control condition. Then maybe to assess outcome interviews? questionnaire? whatever
What are manipulation checks?
Manipulation checks are procedures used in experiments to determine whether the manipulation of the independent variable (levels) was successful—essentially, they verify that the participants perceived and were affected by the manipulation as intended.
could be done before or after experiment!
in other words:
A manipulation check is a way to verify that the levels (treatment conditions) of the independent variable are actually affecting participants as intended, which in turn is necessary for influencing the outcome (dependent variable).
Give some examples of manipulation checks. one before experiment and one after!
Before:
Lets say you wanna set a certain mood in the room cause you think it would affect participants (outcome), you could conduct a pilot study with not that many p. and then assess their moods after with a questionnaire!
After:
Lets say tu eat en place une simulation, debrief after!!! and ask whether they found the simulation to be real or not!!!! ( comme ca tu vas savoir si ta simulation a arm cause le outcome or not) Cause ideally with a simulation, you want it to be as real as possible. EX: zimbardo, standord prison experiment (like your ant the behaviour to be due to the setting/ simulation you know, cause that is your aim to see how they would react in a prison like setting!!!!!)
other example:
placebo control condition. you can ask them after if they believed it to be the true pill!!!!!! bref c important pour savoir si leffet qui y a eu est vrm du a tes manipulations ou a une variable exterieur!!!!!
