Lecture 1 Stats fro Psychology
Lecture 1 Stats fro Psychology 32749
Popular in Statistics fot Psychology
Popular in Psychlogy
This 7 page Class Notes was uploaded by Ashleigh Gaines on Wednesday January 6, 2016. The Class Notes belongs to 32749 at University of Alaska taught by Yasuhiro Ozuru in Spring 2016. Since its upload, it has received 44 views. For similar materials see Statistics fot Psychology in Psychlogy at University of Alaska.
Reviews for Lecture 1 Stats fro Psychology
Please tell me you're going to be posting these awesome notes every week..
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: 01/06/16
January 12 2016 NOTES: Statistics for Psychology A260: Steps In Research Instructor: Yasihuro Ozuru Institution: University of Alaska Anchorage The Ultimate aim of this course: Train you to be a good decision maker when you need to deal with various data/finding that you will encounter in your future as a psychologist. Example: Women are more likely to suffer from depression than men. Statistical issue involved in the above question What kind of data support the above assertion (conclusion)? Steps in Research: Where do statistics fit in the greater scheme of research” Ch.1 4 steps in typical research S1 question and design S2devise measurement S3 data collection S4 data and interpretation Step 1 Research question and design of a study Research question varies in terms of its goal. 1. Merely describing a group of people, animal, etc. 2. Identifying relation between two aspects of variables for a group of people 3. Examining whether a specific causal exists between two variables An Empirical Questionnature of people (or animals) Purely descriptive research Goal To describe or summarize some aspects of a group of people Method Survey Observation Test Example research on description Purely descriptive research Goal What is the number of people in the UAA who suffer from major depression? Method Survey Questions on whether and how people suffer from depression Analysis Compare the appropriate values to answer the question (frequency, percentage?) An Empirical Question relation between variables Associational / Correlational Questions Goal To see if (and how strongly) two or more variables are related Method Correlational design Research on the relation between two variables Association/ Correlational Questions Example Is watching violent TV related to fear of crime? Measure the amount of violent TV viewing Measure the Amount of fear of crime Compute a value(s) that represents the Direction and the Strength of two values (amount of violent TV watching and amount of fear of crime) Difference between association and causality Association/Correlational Questions Advantage: shows if a relationship exists Disadvantage: can’t explain the directions of the relations Questions about causal relation between two variables Causal /Experimental Question Goal * To see if a causal relationship exists between two variables (A&B) Method * Experimental method Standardized procedures for manipulation one variable and measuring its effect (manipulation) or other variable(s) Manipulation by assigning some people to an experimental/treatment group, others to a control/comparison group to isolate the influence of only one variable at a time Experimentation to identity causal relation Causal/Experimental Questions Example: Does watching violent TV cause a change of fear in crime? Analyzing experimental data to answer causal questions IF watching violent TV increases people’s fear of crime THEN fear rating should be higher among experimental participants than control participants. IF watching violent TV does not increase people’s fear of crime THEN fear rating should be similar between the experimental and control participants. Which of the above two hypothesis is true can be examined by comparing the fear rating of the participants – this means you are comparing the two means BUT precisely how much difference (because there always is some difference) is considered to be ACTUALLY DIFFERENT – this requires statistical (probabilistic) decision making Sum on the type of research •Conclusion –Three types of empirical questions •Descriptive •Associational (or Correlational) •Causal (or Experimental) –All types required involves statistics because •Data need to be described •Decision needs to make (e.g., relations, difference) Issues about sampling Across all the three types of research situations, one often do not have access to each of all the individuals whose characteristics (population) you are interested in. Therefore: You need to work with a set of sample. Sample is a subset of population. Sample needs to be representative in order for your conclusion (based on your data) is accurate conclusion about the population Problem 1: Research Design For each of the following scenarios, determine: a) Is the research question descriptive, correlational or causal? b) Whether random selection (of sample) or random assignment is important for the research? c) Whether the research should use a survey, correlational design, or experimental design? d) If an experimental design should be used… I. Which is the independent variable in the study? Ii.Which is the dependent variable? Problem 1: Research Design (cont’d) 1. Interested in the number of REM cycles experienced by adolescents in the US, a sleep specialist has a sample of adolescents spend the night in sleep clinics where their REM cycles are measured. 2. Interested in sex differences in adolescent REM cycles, the same research repeats the above study with two samples: one sample of males and one of females. 3.Interested in possible effects of media on disordered eating, a researcher asks a sample of women to indicate how much time per day they spend reading fashion magazines and how satisfied they are (on a 110 scale) with their bodies. 4.As a followup to the previous study the same researcher assigns female volunteers to sit in one of two waiting rooms for 30 minutes; half sitting in a room stocked with fashion magazines (e.g., Glamour) and half in a room with news magazines (e.g., Newsweek). Afterward, each woman completes a questionnaire about various traits, including their current body satisfaction. Step 2: Devise the Measures •Goal of measurement = Scaling –Translating observations/answers into a series of # that can be analyzed •Different types/levels of measures/scales –Qualitative vs. Quantitative –Four types •Different numerical properties •Different mathematical procedures Step 2: Devise the Measures 1. Choose the category which best describes your family heritage 1. African 2. European 3. Latino 4. Asian 5. Native American 6. Other •Classification Property •When observations can be classified into distinct subgroups •Nominal Variables why is it called NOMINAL? •Scales that merely classify or categorize the observations based on qualitative differences •Minimal mathematical procedures •Qualitative variable 1. Who do you live with? 1. Both parents 2. Mom/stepMom only 3. Dad/stepDad only 4. Sometimes Mom, sometimes Dad 5. Other Step 2: Devise the Measures 1. How far do you plan to go in school? 1. Won’t finish HS 2. HS but no further 3. HS and then tradeschool 4. Some college 5. Finish college 6. Beyond college •Order (Magnitude) Property •When observations can be rankordered from highest to lowest (or vice versa). •Progressive steps •Ordinal Variables •Scales that rank orders observations based on differences in magnitude. •Limited math procedures •Book… ‘Qualitative scale’ •Me… ‘Semiquantitative’ depends….. 1. How many of your close friends smoke cigarettes? 1. None 2. One 3.24 4.510 5. More Step 2: Devise the Measures 1. How often do you eat out? 1. Less than once a week 2. Once a week 3. Twice a week 4. More than twice a week •Equal Interval Property •When any two adjacent #s (e.g., 2 and 3) on a scale are as far apart as any other two adjacent #s (e.g., 7 and 8) •Equally progressive steps •Interval Variables •Scales where the steps on the scale are all equal size. •Quantitative scale •Add, subtract, multiple, divide •True vs. Approximate Interval 1. After smoking for 25 years, how worried should someone be about getting heart or lung disease? 1. Not very 2. A little 3. Moderately 4. Very 5. Extremely Step 2: Devise the Measures 1. What percent of people who smoke for 25 years do you think will get lung cancer or heart disease? •____% (please enter a number from 0 to100) •True (Absolute) Zero •When the scale has a zero point that is not arbitrary •Scores could not possibly go any lower than 0.0 •Ratio Variables •Scales with equal intervals and a true zero •Quantitative variable •Relative comparisons 1. How many cigarettes did you smoke yesterday? • Cigarettes
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'