Chapter 7: Experimental Research Strategy Flashcards
The goal of experimental research strategies are to rule out ______ relationships
coincidental
4 requirements to be a true experiment
1) manipulation: changing the independent variable
2) measurement: a dependent variable is analyzed
3) comparison: scores of one condition will be compared with the scores of another condition
4) control: all extraneous variables are controlled
experimenters are involved in the direct manipulation of the ___ variable
independent
What is a third variable problem?
when a 3rd unidentified variable is CONTROLLING the relationship between 2 experimental variables.
What is an example of a third variable problem?
for example, researchers are determining if there is a relationship between math ability and weight gain in children between the ages of 6-12. They notice that as the children gain more weight, they have increased mathematical ability. However, there is a problem with this relationship: is mathematical ability related to weight gain or is related to maturity as the kids age to 12 and gain more education? Kids will naturally gain weight from age 6-12 so although that is occurring at the same time as their education, their weight gain most likely is not contributing to their mathematical ability.
describe an example where a true relationship is hidden by nature
example: the law of gravity, which states that all objects will fall at the same rate. This relationship is hidden due to air resistance. experimental conditions may test gravity in a no- air environment: will show the relationship
goal of an experiment
to demonstrate the existence of a cause-effect relationship between 2 variables.
What’re the 2 unique elements that completely distinguish an experiment from other research strategies?
1) manipulation of the IV
2) control of extraneous variables.
What is a confounding variable and how does it differ from other extraneous variables?
a confounding variable that is (often unknowingly) allowed to change systematically along with 2 variables being studied, and may provide an alternative explanation for the results seen.
Different than other extraneous variables because:
1) they influence the dependent variable
2) the systematically vary with the independent value
Name 2 problems that manipulation solves in an experimental design and give examples
1) manipulation solves directionality
ex/ it is noted that there is a relationship between ice cream and temperature, but does temperature affect ice cream consumption or does ice cream consumption affect temperature?
you can determine directionality by manipulating the variables: ex/ does increasing the temperature in the stadium lead to more ice cream being eaten during the baseball game? or does handing out free ice cream increase the temperature of the stadium? Most likely the first answer. Therefore, you can determine the directionality of the relationship: temperature affects ice cream.
2) Manipulation solves the third variable problems: you do not need to wait for the variables to change on their own, or else other things may be happening at the same time.
Ex/ increasing temp increases ice cream consumption, but it also increases crime rate. If you allow the temperature to fluctuate naturally/ (ie/ wait for summer), you may notice that crime rate increases and thus may make a conclusion that ice cream consumption is related to crime rate. This is most likely false, and you can prove this by handing out free ice cream and monitoring the crime rates (probably no correlation). You could also create a lot of crime scenarios and see if there is an increase in ice cream consuption (probably unlikely). Therefore, because the variables were manipulated, you can determine that the changes in the independent variable are NOT BEING CAUSED BY SOME OUTSIDE VARIABLE.
3 general categories of extraneous variables
1) environmental (room temp, type of room)
2) Individual differences (age, sex, IQ)
3) time related: treatment conditions are conducted over a period of time–> may cause fatigue
T/F a variable that is changing randomly with the independent variable is a confounding variable
false. A variable is only confounding if it changes SYSTEMATICALLY with the independent variable.
3 Methods of controlling extraneous variables:
1) control by holding a variable constant
2) control by matching values across treatment conditions
3) control by randomization
explain controlling extraneous variables by holding a variable constant and its drawbacks
extraneous variables by holding a variable constant entails making the variable the same for very observation, and standardizing the environment and procedures (ex/ choosing only 6 yo male participants who have completed grade 1, and testing them all in the same environment and same time of day.
A drawback of doing this is that it limits EXTERNAL VALIDITY OF THE EXPERIMENT because there are so many controls, there is also A LOT OF EFFORT INVOLVED
explain controlling extraneous variables by matching values across treatment conditions and its drawbacks.
controlling extraneous variables by matching values across treatment conditions entails allowing the variable to STILL BE PRESENT, but making the variable THE SAME in all treatment conditions. For example; you realize that the group of participants are not uniform in age- 5 are 10 years old and 5 are 15 years old. Instead of scrapping the group and creating uniform group, each separate treatment condition could also have 5 10 year olds and 5 15 year olds- the variability is still there, but the variability is matched.
you could also match time-related factors by varying the ORDER OF TREATMENTS (COUNTER BALANCING) so that you can say that time being tested (earlier rather than later) does not affect results
a drawback of matching variables is that it is time consuming an intensive, usually requiring more measurements to even group participants in the first place.
