How Sociologists Use Statistics

1
How Sociologists Use Statistics
2
History

• First called political arithmetic in 17th
  century England and France and used to calculate
  population size and life expectancy.
• Those who conducted these studies were called
  statists, since the health of the state was
  at issue.
• The first statistics in the U.S. were meant to
  inform debates over social issues in the 1830s
  such as prostitution.

3
History (contd)

• As a result of counting, statists recognized
  patterns peoples actions were more or less
  stable over time.
• During troubled times, the need to reveal these
  patterns were greater than ever before facts
  were needed to guide social policy.

4
Types of Statistics

• --1. Descriptive (e.g., rates, percentages,
  averages)

5
Poverty RatesUS Census 2006-2008

• Milledgeville all families 19.1
• Milledgeville all people 36.2
• Baldwin County all families 14.8
• Baldwin County all people 23.0
• GA all families 11.0
• GA all people 14.5
• US all families 9.6
• US all people 13.2

6
Criticisms of how the US Govt measures poverty

• No adjustments for geographic differences
• Exclusion of taxes and other government benefits
• No way to account for differences in childcare
  and medical care costs
• Inadequate system for adjusting for family size
• (Economic Policy Institute www.epinet.org)

7
Unemployment RatesGeorgia Department of Labor
March 2010

• Baldwin County 14.0 (or greater)
• Georgia 10.4
• National 9.9

8
Criticisms of how the US Govt measures
unemployment

• The formula does not include those who want to
  work full-time but are discouraged and no longer
  looking for work.
• It does not include those who have taken early
  retirement to avoid a layoff, but who prefer to
  be working.

9

• Neither does it include those who are
  underemployed those who have part-time or
  seasonal work but who prefer to be employed
  full-time.
• (wikipedia)

10
Who is counted as unemployed?

• Persons are classified as unemployed if they do
  not have a job, have actively looked for work in
  the prior 4 weeks, and are currently available
  for work. Actively looking for work may consist
  of any of the following activities
• (http//www.bls.gov/cps/cps_htgm.htmunemployed)

11
Overweight/Obesity Rate

• In 1998 public health agencies changed the
  indicators for overweight/obesity.
• Normal 18.5-24.9
• Overweight From 27 to 25-29.9
• Obese 30 or higher
• Yet, in comparisons across time (e.g., from the
  1980s to today), mention is rarely if ever made
  that those indicators have changed.

12
My Criticism

• We are comparing across time while using
  different indicators.
• The Body Mass Index for the overweight/obese
  category has gotten lower, so this means that
  more people are placed in this category without
  gaining any weight.

13
Changing Indicators

• See Food Calories and Land.
• See sociological images website.
• See CDC website.

14
Inferential Statistics

• --2. Inferential Statistics two examples
  confidence interval, rejection of a hypothesis.
• The research well talk about today addresses
  both confidence interval and hypothesis testing
  in order to give us insight on statistical
  significance.

15
Causality

• A hypothesis such as The higher ones Body Mass
  Index (BMI) the less likely one will accept self
  is a causal expression.
• It suggests that self-acceptance is a function of
  or is caused by BMI.
• In inferential statistical terminology that can
  be expressed as the dependent variable
  (self-acceptance) is a function of one or more
  independent variables (BMI).

16

• Sociologists address this question with an
  analysis method called regression.

17

• Statistical significance in sociology is
  indicated with an asterisk. The more significant
  the regression coefficient is, the more asterisks
  that appear by the number. For example
• -.34 represents higher significance than
• -.21 (p. 254).

18
How confident are we?

• Three asterisks mean that there is only a 0.1
  chance that this value represents a non-existent
  relationship.
• This is indicated this way p .001
• As statistics go, this is highly significant.
  The other common p values are
• p .05 (95 confidence interval or CI)
• p .01 (99 CI)

19

• Sowe can reject the null hypothesis that there
  is no statistically significant relationship
  between BMI and self-acceptance.

20
Are there other types of significance?

• Yes substantive significance.
• Substantive significance refers to the
  qualitative importance of the relationship.
• Statistical significance does not necessarily
  imply political, social, or economic
  significance. The relationship found may be so
  small even though statistically significant
  that the variable is of little consequence
  (Schroeder, Sjoquist, and Stephan 1986).

21
Acknowledgements

• Damned Lies and Statistics Untangling Numbers
  from the Media, Politicians, and Activists by
  Joel Best (2001), University of California Press
• Understanding Regression Analysis An
  Introductory Guide by Larry D. Schroeder, David
  L. Sjoquist, and Paula E. Stephan (1986), Sage.