Today

1
Todays Plan

• 430pm530pm elaborating bivariate relationships
• 530pm--630 SPSS(TE lab)
• 630pm700pm dinner break
• 700810pm interpreting data and drawing
  conclusions

2
Elaborating relationships

• Why does the relationship exist?
• What is the nature of the relationship?
• How general is the relationship?
• Elaboration model
• interpretation method
• the Columbia School
• Lazarsfeld method

3
Elaboration paradigm

• Aim
• to provide a logical/statistical
• technique that would allow
• researchers to elaborate on the
• nature of observed relationships

4
The Elaboration Model

• Change in Original
• Relationship After
  Test Factor
• Additional Variable
  Antecedent Intervening
• Introduced
  ---------------- ------------------
• ------------------------------
  
• Same relationship
  Replication
• Weaker or no relationship Explanation
  Interpretation
• Split, one subset the same
  Specification
• or greater, the other is less
• or none

5
How to choose a test factor?

• We choose a test factor that we have reason to
  believe is related to both the independent and
  dependent variables.

6
How do we add a test factor?

• Mechanically, we accomplish this by first
  dividing our sample into subsets on the basis of
  the values of the test variable. The relationship
  between the original two variables is then
  recomputed for each of the subsets. The table
  produced in this manner is called the partial
  table. The relationships found in the partial
  table are called partial relationships.

7
Replication

• Replication occurs whenever the partial
  relationships are essentially the same as the
  original relationship.

8
Replication
Epsilon 12 percentage points
Do you approve or disapprove of the proposition
that men and women should be treated equally in
all regards
9
Replication
Epsilon 12 percentage points
Epsilon 12 percentage points
10
Specification

• Sometimes the elaboration model produces partial
  relationship that differ significantly from each
  other. For example, one partial relationship is
  the same as or stronger than the original
  two-variable relationship and the second partial
  relationship is less than the original and may be
  reduced to zero.

11
Specification (Glock)
Social Class and Church Involvement
Church involvement provides an alternative form
of gratification for people denied gratification
in secular society
People of lower social class have fewer
opportunities to gain self esteem from secular
society
12
Specification
Social Class and Holding Office in Organisations
Social class is strongly related to the
likelihood that a woman has every held an office
in a secular organization
13
Specification
Church Involvement by Social Class and Holding
Secular Office
Mean church involvement for
Rough indicator of gratification in secular
society
14
Explanation

• Explanation is the term that used to describe a
  spurious relationship. The relationship
  disappears when the test factor is included. The
  test factor must be antecedent to both variables.

15
Explanation - spurious relationships

• spurious - not a genuine relationship
• test variable must be antecedent

Strength of peace movement
Likelihood of war
Strength of peace movement
International tensions
Likelihood of war
16
Spurious relationship
Number of fire trucks
Size of the fire
Damage done

• compare the initial bivariate relationship with
  the conditional relationship
• no relationship in the conditional table - we
  have explained the original relationship
• completely and partly spurious relationships

17
Interpretation - Intervening variable (Stoufler)
Education and Acceptance of Being Drafted
Education
Friends Deferred
Attitudes
18
Intervening variable
Relating education to acceptance of being drafted
through the factor of having friends deferred
19
Interpretation

• Interpretation is similar to explanation, except
  for the time placement of the test variable. The
  test variable is antecedent to only the dependent
  variable.

20
Drawing conclusion (Bouma)

• What did you ask?
• What did you find?
• What do you conclude?
• To whom do your conclusions apply?

21
Reporting your research

• the objectives or hypotheses
• the major concept(s) and conceptual definition(s)
• the major variables and operational definitions
• the measurement instrument, for example the
  self-administered questionnaire or content
  analysis sheets

22
Reporting your research

• a description of the research design, for example
  a quantitative case study, an experiment.
• a definition of the population (group of
  interest) and a short discussion of the sampling
  methods you would use in a larger scale study.
  The method for selecting participants for the
  pilot study

23
Reporting your research

• a description of the data collection methods
• the presentation of the results
• a discussion and interpretation of the results,
  i.e. what they mean to you, and any limitations
  or concerns, for example ethical, validity,
  reliability.
• Conclusions