Ryan Fuller

1
Presentation

• By
• Ryan Fuller

2
My Study

• I chose to study the overall sadness of people
  over the age of sixty-five. My main concerns are
  to look at how the effect of a spouse passing,
  bad health, and the presence of children and
  grandchildren effect their overall sadness.

3
Research Questions

• Does being a widow have a positive effect on your
  sadness?
• Is the positive effect of being a widow on your
  sadness weaker when children or grandchildren are
  present?
• How long does the effect of being a widow on
  sadness last?

4
Variables

• Dependant- Overall sadness of members in the
  widow group as well as the control group.
• Independent- Whether or not individual is a widow
  or not
• Control- bad health, sex, number of children,
  number of grandchildren
• Children and grandchildren are also interaction
  variables

5
Statement of Purpose

• Reason this is worth studying?
• Help to understand why the elderly are often sad
• See how long the effect of spousal death lasts
• If the study is proven significant, we could
  present to City and State legislations with
  possible ideas or replacement efforts for widows
  who do not posses any children or grandchildren.

6
Relevant Sociological Theory

• Durkheims Social Integration and Regulation
  Theory
• This theory is relevant because it helps me
  explain why the death of a spouse is an important
  aspect of the persons overall sadness.
• Also, why children and grandchildren may prevent
  sadness when a spouse dies.

7
Quote From Durkheim

• While discussing Suicide, Modern man kills
  himself, Durkheim argues primarily as a result of
  two conditions the loss of cohesion in modern
  society and the absence of suitable moral norms
  by which to orient himself (Zeitlin 287).

8
Explanation of Theory

• Although the subjects in my study are not being
  examined for suicide, this theory explains why
  for instance elderly are often sad. Durkheim
  said that man kills himself as a result of two
  instances, the loss of cohesion in modern society
  is one of them. When a spouse dies these people
  often lose a sense of cohesion and social
  integration which intern will lead to a higher
  level of sadness. This is similar to Durkheims
  idea except we are replacing suicide with sadness.

9
Continuation of Theory

• Core concepts- Integration and Regulation
• Assumptions- society works as a whole to develop
  society. Members of society all play a small
  role in the large functioning body
• Central Proposition- Death of a spouse will
  reduce an individual's integration and regulation.

10
Hypotheses

• Null- Being a widow will not be positively
  related to your overall sadness.
• Alternative- Being a widow will be positively
  related to your overall sadness.
• 2nd Null- the positive effect of widow should
  not be weaker for respondents with children and
  grandchildren.
• 2nd Alternative- the positive effect of widow
  should be weaker for respondents with children
  and grandchildren.

11
Research Design

• I used multiple regression to study the
  respondents.
• They were surveyed at three time periods after
  spouse dies 6 months, 18 months, 48 months.

12
Sampling

• The subjects are residents of Detroit and the
  first wave of the study was in 1987-1988.
• The subjects were selected by using a
  multi-staged approach. First, areas within
  Detroit were studied. Second persons sixty-five
  and older were selected randomly.

13
Data

• These data are available through the
  Inter-University Consortium for Political and
  Social Research.
• Data collected by Prospective Study of
  Bereavement.

14
Strengths of Data

• These data appear to be valid measures of the
  concepts of interest.
• More importantly their face-validity appears to
  be sound.
• Importantly these data allow me to replicate my
  analysis at all three time points.

15
Type of Data

• For purposes of my analysis the variable widow
  (my main independent variable) will be scored 1
  for widows and 0 for the control group of
  non-widows. The main dependent variable is
  sadness, which is measured by the following
  three-category item (saddep) Please tell me
  how often you felt sad during the past week
  (1hardly ever, 2some of the time, and 3most of
  the time).

16
Type of Data Cont.

• My analysis has four control variables. First,
  bad health is measured using the variable badh
  which is based on the following question In
  general, how satisfied are you with your health?
  (1completely satisfied, 2very satisfied,
  3somewhat satisfied, 4not very satisfied, and
  5not at all satisfied). Second, gender of the
  respondent is controlled via the variable
  female that is scored 1 for females and 0 for
  males. Third, the number of children is measured
  using the variable kids that is a simple count
  variable. Fourth, the number of grandchildren is
  measured using the variable gkids which is also
  a simple count variable.

17
Design of Proof

• Testing First Null and Alternative Hypothesis
• I will estimate the effect of the variable
  widow in my model of saddep while controlling
  for the aforementioned control variables. If the
  effect of widow is positive and significant, I
  will reject my first null and let my first
  alternative stand.

18
Design of Proof Cont.

• Testing my 2nd null and alternative hypotheses
• To test my second set of hypotheses, those that
  pertain to the interactions that involve the
  number of children and grandchildren, I will
  create multiplicative terms by multiplying the
  number of children by the widow variable and then
  again by multiplying the number of grandchildren
  by the widow variable. If the coefficients on
  these multiplicative terms are negative and
  significant that means that the positive effect
  of widow on saddep is weaker for respondents
  who have children or grandchildren. Should even
  one such coefficient be found, I will be able to
  reject my second null hypothesis and let my
  second alternative hypothesis stand.

19
Six Months

• Widow .23
• .23
• Bad Health Sadness
• -.23
• Kids

20
Interpretation for Six Months

• Reject first null hypothesis
• Alternative will stand- being a widow will be
  positively related to your sadness.
• Allow my second null hypothesis to stand.

21
Eighteen Months

• Widow .18
• 
  Sadness
• Bad Health .32

22
Interpretation for Eighteen Months

• Reject first null hypothesis
• Alternative will stand- being a widow will be
  positively related to your sadness.
• Accept second null hypothesis
• The positive effect of widow should not be weaker
  for respondents with children and grandchildren.

23
Forty-Eight Months

• Bad Health .20
• 
  Sadness
• Grandkids -.21

24
Interpretation for Forty-Eight Months

• First Null hypothesis will stand
• The effect of being a widow will not be
  positively related to your overall sadness.
• Reject the second Null hypothesis
• Alternative will stand- the positive effect of
  widow should be weaker for respondents with
  children.

25
Limitations

• My use of data for Detroit may be problematic.
  For instance the the City of Detroit mat have a
  positive or negative effect in how people view
  their sadness compared to other cities. As a
  result, the effects uncovered by my study may not
  be generalizable.
• The use of secondary also imposes a limitation
  because the data was not collected with my
  research question in mind.
• When you use panel data you have to allow for
  attrition which is people dropping out of your
  study for whatever reason, which may cause a bias
  in my results.

26
The End