Measurement, Part II

1
Measurement, Part II

• The challenge

2
Levels of measurement

• Nominal Ordinal (nonparametric)
• Interval Ratio (parametric)
• Nominal lowest level, simply classifying
  observations into categories
• Categories should be mutually exclusive and
  exhaustive
• Examples gender, major, religion, state

3
Levels of measurement (continued)

• Numbers assigned to the categories have no
  numerical meaning. Assign individuals, and
  report the falling into each category. Fewer
  statistical techniques can be used
• Ordinal measurement one observation represents
  more of a given variable than another observation

4
Levels (continued)

• Rankings
• Newly developed tests
• Ranks tell whether one observation represents
  more or less than another, but not how much more
  or less--nothing is known about the exact
  difference between any two ranks
• Rankings of crime seriousness

5
Levels (continued)

• Interval like an ordinal scale, but has equal
  intervals between the units of measurement. Not
  only an ordering, but also the same distance or
  degree of difference between observations
• For example, 81 is 1 point away from 80, etc.
• Well-developed tests are interval level

6
Levels (continued)

• With interval measurement, can do addition,
  subtraction, multiplication and division, more
  statistical tests
• Ratio measurement like interval, with the
  additional property of a true zero.
• an individual could have two or three time as
  much of a trait as another with ratio measurement

7
Levels (continued)

• height or weight. A 200 lb person weighs twice
  as much as a 100 pound person
• Not true for interval. For example, no such
  thing as an IQ of 0, and a person with an IQ of
  100 is not twice as smart as someone with an IQ
  of 50

8
Levels (continued)

• Most measurement in the social sciences is
  interval measurement

9
Developing questions

• Most common method of measuring in the social
  sciences is to ask questions
• Types of questions open ended and closed ended
• Open provide own answer. Provides more
  information, may result in ideas or
  considerations the researcher had not tthought
  about

10
Questions (continued)

• Disadvantages of openended harder to categorize
  answers.
• They take longer to answer. Some subjects might
  not answer such questions, and you may end up
  with biased results
• Hite reports. Ann Landers

11
Questions (continued)

• Closed ended questions select an answer from a
  list of choices.
• Advantage quick, easy to code
• Problem making sure all the possible reasons
  are covered

12
Guidelines for asking questions

• Clear, terms should be defined
• Avoid double-barreled (asking two questions in
  one)
• Subjects should be competent to answer the
  questions
• Questions should be relevant to the subjects
• Short

13
Guidelines (continued)

• Avoid negative, or emphasize NOT
• Unbiased
• Recognize social desireability as a factor when
  developing questions (imagine how you would feel
  giving any of the answers)
• Use of contingency questions

14
Guidelines (continued)

• Use of matrix questions
• Try to determine if the ordering of the questions
  will affect the answers
• Rule of thumb if questions are written, start
  with interesting but not threatening questions,
  and put routine questions at the end. With
  interviews, ask routine first so subjects feel
  comfortable.

15
Guidelines (continued)

• Always include instructions
• Pretest questions

16
Sampling

• Population all subjects one is interested in.
  Very large or very small
• Element
• Sample portion of population
• Sampling frame list of people (elements) in the
  populaiton

17
Sampling continued

• Representative sample if the overall
  characteristics of the sample approximate the
  important characteristics of the population
• Biased sample not representative
• Parameters and statistics
• Why sample? time and money

18
Sampling in the U.S.

• Literary Digest polls. Accurate until 1936, when
  Landon was predicted as winner of the
  presidential election
• Reasons (1) low return rates (2 million out of
  10 million) and (2) sampling frame (telephone
  directories and lists of auto owners)
• Poor sampling frames result in bias

19
Sampling in the U.S. (continued)

• 1948 Gallup poll predicted Dewey would win.
  Problems (1) stopped polling in Oct. (2) quota
  sampling
• Two types of sampling probability and
  nonprobability sampling
• Probability sampling uses the laws of
  probability, whereas nonprobability does not

20
Probability

• p number of times an event could occur / total
  number of outcomes.
• Can be express as a fraction, a , as chances out
  of 100, or as a decimal.
• P can range from 0 (no probability to 1
  (certainty)

21
probability (continued)

• A sample will be more likely to be representative
  of a population from which it is selected if all
  members of the population have an equal chance of
  being selected in the sample
• Sampling error error due to the fact that the
  sample is not representative
• Necessity of a complete sampling frame

22
Probability sampling methods

• Simple random sampling (out of a hat, random
  numbers)
• Systematic random sampling every nth element is
  cnosen, select first element at random (random
  start)
• Stratified random sampling
• 1. Divide sample into subgroups based on
  important population characteristics

23
P sampling methods (continued)

• 2. Randomly sample from those subgroups in
  proportion to their percentage in the population
• Choice of stratification variables will often
  depend on what variables are available, and how
  much is known about the population
• This technique most likely to be representative

24
Nonprobability sampling

• Probability sampling only works if there is a
  sampling frame of the population. Sometimes that
  is not possible (i.e., criminals, drug addicts,
  etc.)
• Nonprobability sampling methods, while running
  the risk being unrepresentative might be the
  only option

25
Nonprobability sampling

• Convenience the captive audience
• College students and prisoners
• Purposive researcher uses judgment
• for Example, the mentally ill. Works best if the
  criteria for inclusion are clear.
• Quota like stratified random. Groups are
  selected on the basis of known variables

26
Nonprobability (continued)

• In quota sampling, however, subjects are not
  selected randomly--subjects with the desired
  characteristics are selected until a quota is
  filled for each subgroup
• Snowball each subject is asked to suggest other
  subjects

27
Tips about sampling

• Sample size unusually the number of subjects
  needs to be at least 30. If several groups
  within the sample are to be compared, there needs
  to be at least 10 per group.
• The larger the number of subjects (N), the less
  likely sampling error
• There will always be mortality

28
Tips (continued)

• The greater the heterogeneity of the sample, the
  larger the sample must be. The less population
  diversity, the smaller N might be.
• N is often determined by time and money factors

29
Ethics

• No harm to subjects
• Ethics boards or committees, especially with
  captive populations such as prisoners or children
  (children--parents must give permission
  correctional systems have their own boards to
  protect rights)

30
Ethics (continued)

• Subjects right to privacy
• Confidentiality and anonymity
• The only exception if someone is in danger
• Voluntary participation
• Informed consent nature of the study, possible
  effects, being able to withdraw at any time

31
Ethics (continued)

• Deception and debriefing
• Analysis and report do not fake results, or
  cheat, or conceal technical shortcomings of the
  study
• Milgrams obedience study
• Zimbardos mock prison study
• Tearoom trade study

32
Politics and ethics

• Research can be used for good or for evil
• Ex development of the atom bomb
• Project Camelot assessing internal potential
  for war, actions governments might take

33
Politics and ethics (continued)

• Misinterpretation of studies
• Theory of evolution led to social Darwinism,
  which led to eugenics and justifications for
  Hitlers purges
• Politics may affect how studies are interpreted
  (pornography)