Popular in Introduction to Research Methods in Psychology
verified elite notetaker
Popular in Psychlogy
This 5 page Class Notes was uploaded by Melanie Maino on Thursday March 24, 2016. The Class Notes belongs to PSYC314 at Towson University taught by Brianna Stinebaugh in Spring 2016. Since its upload, it has received 10 views. For similar materials see Introduction to Research Methods in Psychology in Psychlogy at Towson University.
Reviews for PSYC314
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 03/24/16
3/21/16 Chapter 7 (continued) Levels/Scales of Measurement • We can evaluate our Operational Definition by checking… o 1) Reliability o 2) Validity • A good O.D. will be both reliable and valid! • Reliability focuses on consistency and making sure that our measures and procedures are dependable o Use a measurement that has already shown reliability in other studies • 3 Ways to evaluate reliability o 1) Inter-rater reliability: different observers measure the same variable and see how much agreement there is amongst the outcomes/observations. ▯ Correlation Coefficient or reliability coefficient (0.00-‐1.00) ▯ Closure to 1.00 the stronger the agreement is with the outcome and the least likely it will be that you have to change your O.D. ▯ Ex. collect different individuals and fill out questionnaire then come together and measure the reliability of the questionnaire o 2) Test-Retest Reliability: we are checking reliability by comparing scores of people who have completed the same measure two different times. ▯ You can run the same reliability coeffient by using two different score from the same individual if they are highly correlated then you know it worked if they are poorly correlated you must change our O.D. ▯ Ex. give a questionnaire and then in a couple of months give the same questionnaire and you must get the same outcome if you are using test-‐retest reliability. o 3) Interitem Reliability: I am measuring internal consistency amongst the test items. ▯ Split-‐half reliability analysis which splits your measure into two halves and sees how highly correlated they are to one another ▯ Cronbach’s Alpha it is a statistical analysis that picks one specific test item and correlates it with other test items within your measure. If all items are highly correlated then we know we have interitem reliability. • Validity is looking at what we are measuring what we intended to measure. It is the accuracy of the measures we chose. o Manipulation Check: looks at how valid our study is going to be • 5 different ways to perform a Manipulation Check... o Face Validity: picking random people and having them judge how valid our study appears to be (weakest technique) o Content Validity: does the content of our measures fairly reflect the content we set out to study and not any other qualities. ▯ Ex. giving an exam that does not reflect what you originally said was going to be on the exam is low in content validity o Predictive Validity: if my measures are valid and are accurately measuring what I intended to measure then the procedures I choose should produce information that allows us to predict future behavior. Want to help human beings with different psychological disorders and be able to predict the future in terms of those disorders. o Concurrent Validity: comparing scores on a measuring instrument with an outside instrument that has already shown to be valid in measuring that variable you are interested in. ▯ Ex. If I created my own IQ test…if I am trying to see if my new IQ test is valid I give participants new and old instruments of IQ tests and see how they correlate. o Construct Validity: (most important way to measure validity) measuring transition from theory to our current research study. Does my new O.D. for this study reflect overall past research? Should be consistent! • 2 Types of Validity: Internal & External • Overall when you want to test of evaluate a study you look at the Internal Validity and the External Validity • Internal Validity… (3 threats) o 1- Extraneous Variables: any variable or factor that influences the findings of my study that aren’t the independent or dependent variable. If I have a lot of extraneous variables the less controlled my study will be (internal validity will be low). ▯ Ex. exercise vs. memorization…time of day will affect this o 2- Confounding Results: error I get from the value of an extraneous variable that is changing across my different treatment conditions and is not my independent variable. o Extraneous Variables produce Confounding Results o 3- 8 classic threats to internal validity ▯ 1) History threat: this is when some situation or event occurs before my experiment takes place that will influence the results. Any outside events that occurring that effect the results of the study. Ex. fender bender before the study… (it is going to be on your mind) ▯ 2) Maturation threat: any physical or psychological changes within a participant that might effect our dependent variable. • 3 instances… 1-‐ long study (drowsiness, fatigue or boredom influences change) 2-‐ longitudinal studies (cognitive and physical changes over time that may effect the dependent variable) 3-‐ psyc 101/ intro to psyc students (gaining a lot of background and may be able to guess your hypothesis) ▯ 3) Testing threat: threatening effect on our dependent variable from previous administration of the same test. • Pre-‐test/post-‐test design… ▯ 4) Instrumentation Threat: when a part of your measuring instrument changes during the experiment. • Ex. an instrument breaks (computer) • Ex. handwritten: measure each treatment condition equally • Ex. written instruments (survey) language and wording of questions aren’t complex or complicated so when participants receive this instrument they understand what they are dealing with ▯ 5) Statistical Regression Threat: see statistical regression threat occurring when I am grouping participants and assigning them to a treatment condition based on an extreme score of a test • ex. anxiety vs. learning rate: group participants based on level of anxiety…3 groups high anxiety, medium, and low based on scores on the test you received. Coming into a test not knowing what is coming might spike somebody’s anxiety up. ▯ 6) Selection Threat: occurs when you don’t use random assignment. • cannot control individual differences if you are not using random assignment. ▯ 7) Participant Mortality Threat: • 1-‐ participant dropout rate… why are they dropping out of the study? • 2-‐ longitudinal study….looking at a elder and keep an eye out for dementia, cancers, death that may effect your results. ▯ 8) Selection interaction threat: selection paired with any other threat as a combination. 3/24/16 Chapter 8- Solving Problems: controlling extraneous variables • 4 main Types of extraneous variables: o 1-‐ physical variables o 2-‐ social variables o 3-‐ personality variables o 4-‐ context variables o If any are present, internal validity is lowered • 1) physical variables: any physical aspect of a study where the testing conditions have areas that need to be controlled o Ex. time of day, day of week, environment o Ex. testing concentration at 8 am o 3 techniques to control physical variables: *start with elimination, then work your way down to balancing ▯ 1….ELIMINATION ex. if you are studying you want to make sure you are concentrating so you try and eliminate sound by finding a quiet place to study ▯ 2…CONSTANCY – physical variables are consistent within all treatment conditions with ALL participants ex. if you are testing concentration at different times during the day pick randomly from each time slot while making treatment conditions so time is spread out evenly ▯ 3… BALANCING – distribute the effects of an extraneous variable across all treatment conditions within the study • 2) Social variables-‐ qualities of a relationship between the participant and the researcher that can influence the results of the study o 2 main social variables ▯ 1-‐ demand characteristics: aspect of study that demands participants to act in a specific way • participants usually conform to how they think they should act • being a participant automatically alters behavior, which produces confounding results • participants may even try to guess what you are studying • 2 ways to control demand characteristics: o 1…single-‐blind studies-‐ participants do not know what treatment they’re receiving ▯ prevents them from guessing hypothesis ▯ ex…drug studies…do not know if they are getting medicine or placebo o 2…cover story-‐ false explanation of procedures being used to keep participants from guessing hypothesis ▯ using deception-‐ must DEBRIEF participants • 2-‐ experimenter bias-‐ an experimenter can give participant cues for how they want them to respond without realizing it o human and animal studies o Rosenthal effect-‐ experimenter treats participants differently, depending on what is expected of them ▯ Ex. giving the treatment condition who is receiving medicine more time that the placebo treatment condition o Experimenter becomes the extraneous variable o Can be seen when experimenter makes an error recording notes o How to control experimenter bias ▯ Double-‐blind study-‐ BOTH the participant AND experimenter do not know who is receiving which treatment • Controls for demand characteristics and experimenter bias • Use whenever possible because it controls the most • 3) personality variables: any personal characteristic that an experimenter brings to the study o warm and friendly experimenter get positive responses ▯ less dropouts ▯ more compliance ▯ better results o Hostile/aggressive experimenters get the opposite o Have to be consistent among treatment conditions o Ways to control personality variables: ▯ 1-‐ consistent as possible • write a script to follow ▯ 2-‐ use multiple experimenters • same number of participants within each treatment conditions ▯ 3-‐ minimize face to face contact with participants o To control participant personality variables use random assignment ▯ Every personality has an equal chance of being in each treatment condition ▯ Avoids selection threat
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'