Conducting Psychology Research PDF

Conducting Psychology Research  Research Goals Internal Validity - The degree to which a cause-effect relationship between two variables has been unambiguously established Causati on True experiments include… Independent Variable (IV)...

Conducting Psychology Research  Research Goals Internal Validity - The degree to which a cause-effect relationship between two variables has been unambiguously established Causati on True experiments include… Independent Variable (IV): The factor that is manipulated by the experimenter Dependent Variable (DV): The factor which is measured in the experiment It depends on the IV Constants: All the factors that stay the same in an experiment Imagine you want to know which color birds prefer What is the possible cause-effect  Our Experimental Design Red Blue Green Independent Variable Dependent Amount of Seed Eaten Variable Constants Location of feeders, kind of seed, type of feeder  Research Goals Internal Validity - The degree to which a cause-effect relationship between two variables has been unambiguously established Is this research design high in internal validity? Yes! Any changes in seed eaten should be caused by the color differences  Sam’s Experimental Design Sam wants to know if birds have color preferences. He makes 3 identical feeders and paints 1 red, 1 blue, and 1 green. He fills the feeders with the same amount of sunflower seed and plans to keep the feeders out for 1 week before measuring how much seed is eaten out of each. On Tuesday, he puts each of the feeders in his backyard: the red feeder in a large dead tree the blue feeder on the doghouse the green one in a small bushy tree Sam’s Experimental Design Red Blue Green Independent Variable Dependent Amount of Seed Eaten Variable Constants Kind of seed, type of feeder, amount of time outside Other Variable  Research Goals Internal Validity - The degree to which a cause-effect relationship between two variables has been unambiguously established Is this research design high in internal validity? No! Cause = color OR location!  Threats to Internal Validity Confounding variables - Other variables that change with the independent variable For high internal validity, we must eliminate confounds  Maria’s Experimental Design Maria wants to know if birds prefer one color of feeder to another. She makes 3 identical feeders and paints 1 red, 1 blue, and 1 green. She fills each of the feeders with the same amount of mixed birdseed and plans to keep the feeders out for 10 days before measuring how much seed is eaten from each feeder. She puts each of the feeders in the school courtyard in a small tree, hanging each at the same height. Maria’s Experimental Design Red Blue Green Independent Variable Does this design seem Dependent Amount to be of Seed Eaten Variable high in internal validity? Height of feeders, location of feeders, Constants kind of seed, type of feeder, amount of time outside  Jessie’s Experimental Design Jessie wants to know if birds prefer one color of feeder to another. She makes 3 identical feeders and paints 1 red, 1 blue, and 1 green. She fills the each of the feeders with sunflower seed. She plans to leave each feeder out for 2 days and then measure the amount of seed that the birds ate. She has one hook in her backyard that she plans to hang each of the feeders on. She puts the red feeder out on Tuesday and measures the seed 2 days later, the green feeder doesn’t go out until Saturday since it was very snowy, but she is able to measure the seed 2 days later. The weather turned rainy, Jessie’s Experimental Design Red Blue Green Independent Variable DependentDoes this design seem Amount to be of Seed Eaten Variable high in internal validity? Height of feeders, location of feeders, kind of Constants seed, type of feeder, amount of time outside Other Variable  Threats to Internal Validity Confounding variables - Other variables that change with the independent variable For high internal validity, we must eliminate confounds If I drink Mountain Dew before bed, then I will not sleep very CAUS much. E EFFEC T IV being manipulated: Mountain Dew vs. no Mountain Dew If I leave all the lights on all day, then my electric bill will be expensive IV being manipulated: Lights on all day vs. lights off all day DV being measured: If I brush my cat more, then there will be less fur on my furniture IV being manipulated: Brush my cat more vs. brush my cat less DV being measured: Amount of fur on the furniture Elizabeth wanted to test if temperature affected how fast milk goes bad and curdles. She left milk in a closet at room temperature, a fridge, and an oven that was turned on low heat. She then measured how soon the milk showed signs of going bad. IV being manipulated: What are some Temperature potential confounds? DV being measured: How soon the milk curdles  Threats to Internal Validity: Mrs. Smith M R S S M I T H Regression Selection Selection Instrument History Maturation to the by Mortality Testing of Subjects ation Mean Maturation Although Changes in Bias in Extreme groups were test scores Subjects selecting & scores similar at may be Extraneous change assigning Dropping Changes in become less first, they because of events between participants out of the measuremen extreme on have now practice have test to groups study t procedures subsequent developed from occurred sessions results in dif. tests differently repeated b/t groups over time tests Do children like sweets more than adults do?  Rashi’s Experimental Design Rashi, a graduate student in psychology, was interested in exploring whether children like sweets more than adults do. He brings 10 boys and 10 girls into the lab and lets them eat marshmallows. 85% of the children opt to eat them. He brings 10 men and 10 women into the lab and repeats the study. Only 20% of the adults opt to eat the marshmallows.  Rashi’s Experimental Design Independent Variable Dependent Amount Does this design seem of Variable Marshmallows to be Eaten high in internal validity? Length of time to eat, type of marshmallow, amount of Constants marshmallows available, location of marshmallows Other Variables…  Threats to Internal Validity Confounding variables - People of different ages differ in many other ways, too Suppose you want to know if diet causes changes in sleep patterns Which diet would you rather eat? Suppose you want to know if diet causes changes in sleep patterns Confounds ?  Threats to Internal Validity Confounding variables – What else do these people have in common? Could another commonality be the true cause? SOLUTION: Random Participan t Sample Group 1 m o e d n nm a R ig s s t A n Group 2  Random Assignment The experimenter randomly assigns participants to groups Participants can’t choose which group they are assigned to Group 1 Group 2 How does this affect internal validity?  Random Assignment Not always possible Age differences Gender differences Personality differences Political affiliation Others? “Experimentation in the laboratory occurs, socially speaking, on an island quite isolated from the life of society.”  Limitations of Experimental Research? Ethical ? Practical/Logical? Realistic?  Would theseResearch Goals results matter Internal in realValidity life? - The degree to which a cause-effect relationship between two variables has been unambiguously established External Validity - The degree to which a study ensures that potential findings apply to settings and samples other than the ones being studied. Evaluating internal and external validity