General Slides for SOC120 Fall 2005 Week 9,10: C11, C12 edited Study Guide edited 30607

1
General Slides for SOC120Fall 2005Week 9,10
C11, C12 (edited Study Guide edited 3/06/07)
2
Study Guide C10 8th Ed

• DiffirenceInformal and Formal Inductive
  Arguments
• (1) first level Everyday life (340b)
• (2) second level Scientific (340b)
• Obvious diff in 1 and 2 (340b)
• EXAMPLE OF FORMAL Political Poll (341)
• RANDOM VARIATION PROBLEMS AND SOLUTIONS
• (1) Error margin (p343t)(population variability)
• (2) Confidence level/SIGNIFICANCE (p343m)
• Rule IV (p343m)
• G. SAMPLE SIZE (p343)
• (1) Short Answer (p343b)
• (2) Approximate error Margins (BOX 345)
• (3) Typical confidencelevel Significance(p345b)
• H. INFORMAL INDUCTIVE ARGUMENTS
• (1) Cautious conclusions EX (p346)
• (2) Summing Up Steps 1, 2, 3 (p347)
• I. FALLACIES (p349t)
• (1) hasty generalization/overgeneralization
• (a) Anecdotal evidence

• Inductive Argument
• Def/Ex (p334t)
• Probability (p334t)
• Strong ---Weak
• Ex need for prediction (p334m)
• 3 examples (p335t)
• Motorcycles
• Barking dog
• Peaches
• Analogical Arguments (p335b)
• Def/Ex
• terms
• target
• target/feature/propety in question
• Rule 1 (p336t)
• Peculiarity of arguments (p336b)
• Inductive Generalizations
• 1. Def/Ex (p337b)
• 2. Sample (p338t)

3
Confidence/Significance and Error margin
Guess your weight within
3 lbs
Hypothesis Peoples Attitude about
abortion is related to their religious identity.
Chisq(P) 61.63(p 0.00

• The researcher knows that
• His/her guess is the hypothesis
• The significance was decided before the study at
  95 out of 100 guesses, .05. Confidence/significa
  nce
• RELIG and ABANY are related since the Chisq was
  .00 and the significance at the .05 level is
  accepted

• The guess for my weight
• His/her guess will be X (151 lbs)
• The guess will have to be 3 lbs of X, 148-154
  lbs Error margin
• He/she can do this 95 out of 100 guesses
  Confidence/significance

4
C11 Terms Study Outline 8th Ed
BOX Couch Potato p382 (d) Size of
group and statistical significance chart
p383t (e) Do not assume always check 1,2
p382b (f) Even if significant first,
second, 3 p383b BOX Cigarettes,
Cancerp384t (2) Nonexperimental cause to
effect def/ex p384 slide
(a) Do not assume1-4 p385t BOX HPV
vaccine found p386t Self Select vs Match
p385b (3) Non Experimental Effect to
Causeslidep386b D Appeal to Anecdotal
Evidence Def/Ex p387b (1) To establish X as
causalp388m BOX The Farside p389t
BOX The Wrong Initialsp388t E
Doubtful Causal1-5 p388-390 F Causal
explanation Arguments p390-391 (1) The moral
is obvious..p391t The lesson is
simple..p391b (2) Explanations and excuses
p391-393 BOX Study Backs Old Idea
p392 FALLACY confusing explanations
and excuses p393

• A (1) Intro Causal Arguments
• Common example Def/Ex
• Post hoc fallacy Def/Ex
• Correlation Cause p371
• (2) Informal causal reasoning
• Causal claim/cause-and effect-claim/Hypothesis
  def/ex p372m
• (3) Two Basic patterns of causal reasoning
• (a) Relevant differencep373mExample BOX
  p374t, Sum p376m
• (b) Common-thread reasoning p374 Sum p397b
• (c) CT best for?, RD best for? 376b
• B Common Mistakes In Informal Causal Reasoning
  1-5 p377
• BOX Converting BOX p374t
• BOX The Great 9/11p375t
• BOX Televisions effectp377
• C General Causal Claims Def/Ex p379b
• (a) Controlled cause to effect Def/Ex p380
  slide
• (b) Concepts p 381m
• (c)Statistical significance p381b-Me

5
Table Reading for A05 Part A

• Table conventions Creating and Reading a Table

Reading a table involves the following (1) Big
Picture General characteristics of the table.
What is the (a) Population and sample (size for
sample, acceptable?) (b) Independent variable
categories and distribution (acceptable?) (b)
Dependent variable categories and distribution
(acceptable?) (2) Specifics distributions of
interest within the table (3) Conclusion
hypothesissupport, reject, suspend judgment (see
Part B) (4) Problems needs for further study,
variable categories, size and distributions of
categories
6
Table Reading for A05 Part B

• Table conventions Creating and Reading a Table
  (interpreting the distributions in a table for
  your hypothesis)

If your independent variable has two many
categories to run across a page, simply make it
the row variable and calculate(row) and compare
percents across column --gt next page
7
Simple Table
Hypothesis males in the US have a higher average
income then females
If the hypothesis is supported which cells A, B,
C, D, should have the highest percents and which
the lowest? (A and D)
8
Tables
V3 AGE by V46 CONCERNED ABOUT CONTRACTING
AIDS V46
Page 1 of 1 Count Row Pct
very somewhat not very not conc
erned Row
1 2 3 4
Total V3 --------------------------------
-------- 1 283 207
115 62 667 19 or younger 42.4
31.0 17.2 9.3 33.2
--------------------------------
2 216 119 83 34 452
20- 21 47.8 26.3 18.4 7.5
22.5 -----------------------
--------- 3 209 163
91 41 504 22- 24 41.5
32.3 18.1 8.1 25.1
--------------------------------
4 89 63 28 26 206
25- 29 43.2 30.6 13.6 12.6
10.3 -----------------------
--------- 5 39 30
32 13 114 30- 39 34.2
26.3 28.1 11.4 5.7
--------------------------------
6 14 14 13 10 51
40- 49 27.5 27.5 25.5 19.6
2.5 -----------------------
--------- 7 4 3
4 2 13 50 or over 30.8
23.1 30.8 15.4 .6
--------------------------------
Column 854 599 366 188
2007 Total 42.6 29.8 18.2
9.4 100.0
Across Read down if reverse Indep and Dep
Can have Row as Independent Col.. as
Dependant by row Read by comparing across
dependant (col.. this ex)
Chi-Square 35.10558 DF
18 significance .00917
9
Ideal Experimental Modelcontrolled cause to
effect
How much difference? See significance
Experimental group factor X is introduced all
else stays same Control group all stays the
same C--- the suspected causal agent (c in the
graphic above) E ---the effect being
investigated
?C11 Concepts
10
Non experimental cause to effect two ex
Difference from experimental cause to effect is
the experimental agent is not induced by
researcher due to ethics, laws, money, etc.
A. Find a group I who have been exposed to
causal agent rather then investigator exposing to
causal agent and create a matching group "II".
Examine change over time for both groups. (e.g.
Study of nurses and HMO patients life, diet, etc.
patterns
B. Find a group I who have been exposed to
causal agent rather then investigator exposing to
causal agent and examine change after a period of
time I to II. (e.g. Study of Beagles that over
time naturally develop skin cancer. It would seem
as if the "causal agent" caused the cancer shows
importance of group III and IV.)
?C11 Concepts
11
Non experimental effect to cause
Find group II that has condition and compare
with a group, "IV" that doesnt have the
condition find difference (e.g. Mouth cancer
determine difference is smoking cigars Many
possible differences in II and IV would limit
strength of conclusions
?C11 Concepts
12
C10, C11

• Chapter 10
• Representativeness and Bias,
• Sample size,
• Error Margin
• Confidence Level
• Criteria and Fallacies of Inductive
  Generalization
• Chapter 12
• Classifies research in terms of pattern of causal
  reasoning, models of research and fallacies.

13
Significance/Confidence

• Probability/possibility that an event could have
  occurred by chance
• For tables use the ChiSq value
• .05?means this distribution could have occurred
  by chance in only 5 of 100 times if the study
  were repeated 100 times. This means you are
  confident that you are confident of the accuracy
  at the .05 level.
• Typically social science operates at the .05
  level. If the number were larger .06, .08, .1
  it is not considered significant, we do not have
  confidence that a relationship found in the table
  distribution is real

Polls and surveys S6
14
Ambiguous Student Research Statements from A05

• A hypothesis concerning this theory is whether
  gender is affected by the opinion of stronger gun
  laws
• I chose to research the topic of marital status
  in regards to the abortion decision because I am
  interested in how a womens companionship with
  others may affect her belief that abortion is an
  individual matter.
• I believe that illegal drug use is spread among
  all ages but, I am concerned in the amounts among
  teenagers and adults using them.
• This paper focuses on the effects of education,
  as they pertain to the satisfaction level of an
  individual with respect to state government
  performance.
• Based on a mothers ethnicity, we can accumulate
  an idea on what level of education she possesses
  or will possess.
• For this research project I would like to
  determine the relationship between Capital
  Punishment and the right to choice about abortion
• The topic I have formulated is Gender Roles and
  their views regarding Capital Punishment.