Chapter 2 Research in Psychology

Modified: 2006-02-17

• AN EXAMPLE OF PSYCHOLOGICAL RESEARCH
• CATEGORIES OF RESEARCH IN PSYCHOLOGY
  • Experimental Research
    • Scientific research that compares two or more groups that received different treatments.
  • Correlational Research
    • Scientific research that predicts the outcome of one variable based on the scores of one or more other variables.
  • Meta-analytic Research
    • A quantitative statistical technique for combining the results of several studies.
• AN OVERVIEW OF THE RESEARCH PROCESS
  • 1. The Work of Others
    • Re-searching what others have already done
  • 2. The Idea
    • The creative part of science
  • 3. Design
    • Asking the right questions
  • 4. Ethical Issues
    • Asking the questions in a proper way
  • 5. Logistics
    • All the nitty-gritty details
  • 6. Collect Data
    • Applying the design
  • 7. Analyze Data
    • Seeing what you found
  • 8. Conclusions
    • Making sense of what you found
  • 9. Present Research
    • Telling others
• EXPERIMENTAL RESEARCH IN PSYCHOLOGY
  • Scientific research that compares two or more groups that received different treatments.
• TWO EXAMPLE EXPERIMENTS
  • Motivation and Performance in Sports
    • Putting by motivated and unmotivated students
  • Music Tempo and Productivity
    • Problem solving under changing tempos of music
• ELEMENTS OF EXPERIMENTAL RESEARCH
  • The Generalized Experiment
  • Independent Variables and Dependent Variables
    • independent variable
      • A variable whose values are controlled by the researcher; the cause in cause-and-effect.
    • dependent variable
      • The variable that is expected to change as a result of changes in the IV; the effect in cause-and-effect.
  • Levels of the Independent Variable
    • Values of the IV that the researcher chooses for an experiment.
  • Nuisance Variables
    • A factor that causes the DV scores to be more variable.
  • Confounded Variables
    • A variable whose levels change in step with changes in the IV.
  • Controlled Variables
    • An extraneous variable that does not vary concomitantly with the IV.
  • Extraneous Variables
    • General term for a variable that varies concomitantly with the IV.
• POPULATIONS AND SAMPLES
  • Random Samples or Probability Sample
    • A sample from a population obtained by applying the random sampling technique
      • Represent mathematically-based subsets of a population and are typically used in quantitative research.
    • Simple Random Sampling
      • A procedure is established to create a random sample from a population. For example, pulling 100 names (the sample) from a container of 1000 names (the population).
    • Systematic Random Sampling
      • A random sample is created by choosing the nth person from a large ordered record of a population. The start point of the selection process must be randomly selected. For example, making a sample from a telephone book by selecting every 64th person starting at a random page and location and then running through the entire phone book. Typically, the interval is selected by dividing the size of the population by the size of the sample desired.
    • Stratified Random Sampling
      • When equal or nearly equal numbers of participants are desired along certain categories or dimensions, a stratified random sample is used. The groups, or strata, are selected ahead of time, and then random samples are created by one of the techniques above. For example, if two random samples, one for males and one for females are desired, then two collections of names are created. One collection contains the names of all of the males in the population the other contains the names of all the females in the population. Random samples are then created from each sub-population.
  • Convenience Samples
    • A sample that was not obtained randomly from the population.
    • Non-Probability Samples
      • Represent samples that do not rely or conform to mathematically-based subsets of a population. Instead, they are created quasi-randomly, conveniently, or in an ad hoc manner. External validity is usually lower when non-probability samples are used. These types of samples are used in both quantitative and qualitative research.
    • Convenience, Accidental, or Availability Sampling
      • These multiple terms are used to describe samples created easily. Much social science research comes from these kinds of samples. For example, using participants from college classes is a very common form of this kind of sampling. However, researchers should realize that external validity may not be very high with this kind of sample.
    • Community Snowball Sampling
      • Like the fist size snowball that when rolled becomes the large lower body of the snowman, this sample grows. In this technique, a small initial sample is recruited, and then those participants, in turn, recruit additional participants similar to themselves. For example, an initial small group of drug users could be asked to find other drug users like themselves.
    • Judgmental or Purposive Sampling
      • In this sampling technique, researchers select participants based on preselected criteria. Typically, judgmental or purposive sampling is used to study groups not well represented in the population. For example, parents who home school their children might be used as a sample.
    • Quota Sampling
      • In quota sampling, a sample is created that approximates previously known characteristics of a population. However, the matching procedure is not mathematical nor based on sampling theory. Instead, rough approximations are made of the population's characteristics and then a sample is created to match those them. For example, a researcher might create a sample that roughly matches the age distribution of the population it is drawn from.
• INTERNAL VALIDITY AND EXTERNAL VALIDITY
  • internal validity
    • Degree to which extraneous variables do not affect dependent variable scores.
  • external validity
    • Ability of a study to generalize to other populations, situations, or conditions.
  • relationship if internal and external
    • internal is more important
• DESIGN AND STATISTICS
  • design
    • The plan for Collecting data.
• STATISTICAL METHODS
  • Descriptive Statistics
    • Numbers or graphs that summarize a data set.
      • mean, median, mode
      • range, standard deviation, variance
  • Data Exploration and Description
    • visual examination
    • graphs
      • line, bar, pie charts, others
  • Correlation Coefficient
    • Pearson r
      • Range: +1.00 to -1.00
      • two or more variables
      • shows degree of relationship AND direction of relationship
        • height and weight, weight and price
  • Inferential Statistics
    • A method of reaching conclusions that uses samples and probability.
    • chi square, t-test, ANOVA
• PUBLICATION AND PERSUASION
  • A Seven-Point Outline (APA) Prospectus Items in Bold
    • 1. Title
    • 2. Abstract
    • 3. Introduction
    • 4. Method
    • 5. Results (Expected Results)
    • 6. Discussion
    • 7. References
  • Publication
    • Part of science
    • Types
      • Oral
        • informal, meetings
      • Written
        • self-published, published by others, peer review
  • Persuasion
    • Data do NOT speak for themselves
    • Scientists persuade other scientists

IN THE KNOW

In The Know--In 1896 Karl Pearson introduced the correlation coefficient that today bears his name. In the more than 100 years since its introduction, Pearson's versatile statistic has been expanded and applied in many, many fields.

In The Know--There is a moderate revolution underway among researchers regarding what inferential statistics to use to analyze data from experiments. In this book we put most of our emphasis on the traditional NHST approach, but we describe the controversy and also illustrate some of the techniques advocated by the revolutionaries. The goal of the reformers is to have more information than that yielded by a NHST test. In particular, they advocate statistics that yield more than the greater than or less than conclusion provided by NHST tests.

In The Know--In the old days when an author sent material to an editor, manuscript was an accurate word; the report was written by hand with a pen. When typewriters came into use around the turn 20th century, the term typescript didn't catch on and manuscript continued to be used. At the turn of the 21st century, material to be published is submitted as a word-processed, electronic file. So far, we still refer to the file as a manuscript.

GLOSSARY

• confederate
  • A member of the research team who appears to be a naïve participant but instead acts a pre-determined role.
• confounded variable
  • A variable whose levels change in step with changes in the IV.
• control group
  • The generic name of the group in an experiment that does not receive a treatment.
• controlled variable
  • An extraneous variable that does not vary concomitantly with the IV.
• convenience sample
  • A sample that was not obtained randomly from the population.
• correlational research
  • Scientific research that predicts the outcome of one variable based on the scores of one or more other variables.
• dependent variable
  • The variable that is expected to change as a result of changes in the IV; the effect in cause-and-effect.
• descriptive statistics
  • Numbers or graphs that summarize a data set.
• design
  • The plan for Collecting data.
• experimental group
  • The generic name of the group in an experiment that receives a treatment.
• experimental research
  • Scientific research that compares two or more groups that received different treatments.
• external validity
  • Ability of a study to generalize to other populations, situations, or conditions.
• extraneous variable
  • General term for a variable that varies concomitantly with the IV.
• independent variable
  • A variable whose values are controlled by the researcher; the cause in cause-and-effect.
• inferential statistics
  • A method of reaching conclusions that uses samples and probability.
• internal validity
  • Degree to which extraneous variables do not affect dependent variable scores.
• levels
  • Values of the IV that the researcher chooses for an experiment.
• logistics
  • The materials needed and the plans for carrying out an experiment.
• manuscript
  • A typewritten report or an electronic file of a research project.
• meta-analysis
  • A quantitative statistical technique for combining the results of several studies.
• nuisance variable
  • A factor that causes the DV scores to be more variable.
• null hypothesis statistical testing
  • A statistical technique that produces accurate probabilities when the null hypothesis is true.
• operational definition
  • Procedures or operations used to measure a variable or establish a condition.
• peer review
  • The process by which other scientists judge the value of completed research.
• population
  • An entire set of scores.
• random sample
  • A sample from a population obtained by applying the random sampling technique
• sample
  • A subset of the population
• scatterplot
  • A graph of participants' scores on two variables.
• social loafing
  • When a person works less as a group member than as an individual.
• treatment
  • Values of the IV that the researcher chooses for an experiment.