Statistics

1
StatisticsChapter 4

• Analyzing Frequency Distributions
• Read pp. 89-95, 99-105, 116, 118-119, 121-122

2
Frequency Distributions

• Univariateexamining one variable at a time
• Raw data/scores data that has not been processed
  and scores that have not been summarized in any
  way.
• Frequency distribution a summary of the
  responses to the categories of a variable
• In its simplest form, it lists the frequencies of
  response (number of responses) to each category
• Percentages of response to each category

3
Frequency Distributions

• Sometimes, the researcher will not report certain
  responses
• In other cases, an abbreviated version of the
  variable is created by combining several
  responses. This is called recoding

4
Characteristics of Frequency Distributions

• They each describe a variable of interest
• They list the categories of a variable
• They list the distribution of responses for the
  categories the variable

5
Terminology 2

• Statistics tells valid cases for the variable
  i.e., number of meaningful responses to a
  questions
• Missing cases number of cases for which
  respondents did not know the answer to q
  question, did not answer the question, or were
  not asked a the question for a variable
• N number of valid cases

6
Percent and Valid Percent 3

• Percent Frequency of response to a particular
  category divided by the sum of all cases, valid
  and missing
• Valid Percent Frequency of responses of a
  particular category divided by the sum of valid
  cases only
• Cumulative percent running total of all
  responses, computed by adding the number of
  responses to those of all preceding categories,
  divided by the number of valid cases (see p. 100,
  table 4.1)

7
Interpreting the frequency distribution
4

• Heterogeneous distributions those in which the
  respondents are fairly evenly distributed, or
  spread out across all categories of the variable
• Homogeneous distributions those in which the
  respondents are clustered or grouped into only a
  few categories of a variable
• Pitfallsp. 103
• Split filedividing up the data file by the
  characteristics of the respondents

8
Skills practice 1, p.105

• Produce frequency distributions for marital,
  degree, wrkstat, and xmarsex

9
Recoding and splitting 2

• Recoding Allows us to change how the respondents
  to a variable are grouped together to simplify
  the analysis
• Recombining values into variables with fewer
  categories makes it easier to see patterns
• Split filedividing up the data file by the
  characteristics of the respondents
• Can see how a particular age or gender group
  answered a particular question

10
Homework

• All skills practices in the above sections (not
  to be handed in)
• P. 151/ 1-5 (hand in appropriate graphs for
  1,2,5)
• SPSS p. 154/ 1,2,3,11,12,13 (hand in 12)