LESSON 5

Dr. John M. Ritz

CHAPTER III

    * METHODS AND PROCEDURES Introduction (research type and sub-sections to be described)
    * Population
    * Research Variables (experimental studies)
    * Instrument Design or Use
    * Field, Classroom or Lab Procedures (experimental studies)
    * Methods of Data Collection
    * Statistical Analysis
    * Summary

Sampling Process

    * Select a sample that represents a group.
    * Gather data from the sample.
    * Use the data to make references about a larger group, the population or universe.
    * Make decisions based upon what the sample told us.

Sampling Techniques

    * Population (all of a select group)
    * Sample (select group or finite set of people, objects, or things taken from the whole)
    * Random Sample (typical sample, representative of the whole)
    * Stratified Random Sample (sub-populations or strata included to give a truer picture of the whole population, guarantees sub-group representation)

Small vs Large Samples

    * Small - convenient and economical, however there is a greater chance of error.
    * Large - more accurate, however more costly and time consuming.

Validity

    * relates to instrument design It deals with whether the instrument is appropriate to the particular problem being investigated.
    * The instrument must be relevant to the particular questions being asked, not merely related questions.
    * It is concerned with the kind of information a score, rating or evaluation yields about an individual's behavior in a particular setting.
    * A measurement instrument is said to be valid if it measures what it is supposed to measure.
    * Does it do what it is supposed to do!

Types of Validity

    * Construct Validity - related to characteristics that are believed to account for some aspect of behavior - theory. It is based on a theory of what you think the concept is (what really is intelligence, leadership, etc.).
    * Content Validity - the adequacy that an instrument samples a given situation. After construct validity is established, identify the content that encompasses the construct (problem solving, IQ, mathematics ability, etc.).
    * Criterion-Related Validity - relating the results of one instrument to the results of another test (IQ tests or math or reading tests).

Internal Validity

    * focus on the entire experimental design Internal validity in an experimental design asks the question: "did, in fact, the experimental treatments make a difference in this specific instance?"
    * Did the independent variable X really produce a change in the dependent variable Y?
    * Must control for extraneous variables such as:
          o History
          o Maturation
          o Testing
          o Instrumentation
          o Statistical Regression
          o Selection
          o Experimental Mortality
          o Selection-Maturation Interaction

External Validity

    * focuses on the entire experimental design External validity asks the question: "to what populations, settings, treatment variables and measurement variables can this effect be generalized?"
    * What relevance do the findings concerning the effect of X have beyond the confines of the experiment?
    * Must control for extraneous variables found in:
          o Selection biases and external variables.
          o Effects of pre-testing.
          o Experimental procedures.
          o Multi-treatment interferences.

Reliability

    * focuses on instrument design Does the instrument produce the same results consistently (consistency).
    * Methods for testing for reliability:
          o Test-retest method
          o Parallel forms method (equilivant forms method)
          o Internal consistency form (split-halves method)