Agenda

1
Agenda

• Observational research

2
Beginning steps

• 1. Determine research question(s)
• 2. Determine hypotheses
• 3. Narrow behavioral categories to those which
  answer the research questions.

3

• Make informal observation of participants.
• Be familiar with the behaviors exhibited by your
  participants.
• (e.g., go to a volleyball game and watch and note
  behaviors)
• Write your observation categories
• Construct a complete list.
• Condense to fewer categories later

4

• From literature search -- how did other
  researchers define behavioral categories.
• If use same categories, its easier to compare
  your results with those previously reported.

5

• Decide on coding schemes
• Define general and specific subcategories
• Define what behaviors will be recorded during
  observation time

6

• Frequency (within a set time period)
• Duration (within a set time period)
• Interval Divide observation time into discrete
  time intervals, then record whether a behavior
  occurs within each interval.

7
You cant observe everyone -- Sample

• Be realistic
• Sample rather than observe and record every
  occurrence.
• Sample by time (period of observation)
• Sample by particular participants -
• limit of people observed for __ minutes and
  record only their behavior.
• Sample by situation (different settings/
  conditions for observations)

8
Benefits of videotaping

• Slow down, rewind tape,
• Multiple observers
• Review several times and see something at later
  time that you didnt see at first.
• Code later

9
Sources of observer bias

• Knowing the goals of study or hypotheses being
  tested may influence their observations
• Example. Coding of male child roughly taking a
  toy away from another child different than female
  child doing same thing because hypotheses is
  males show more interpersonal aggression.
• Strategy Use blind observer - unaware of
  hypotheses.

10
Observer bias

• When observer EXPECTS certain behaviors or events
  to occur. (e.g.Zechmeister exercise)
• When interprets rather than simply record
  behavior.
• Inferences may or may not be true.
• Strategy Record behaviors rather than
  interpretations of behaviors.

11
Personal biases and expectations

• May affect how you observe and record behavior.
• If you were only 1 observer, couldnt establish
  the reliability of your observations.
• Strategy Use 2 or more trained observers.

12
Multiple observers

• It is unlikely that multiple observers will
  always agree.
• They will differ in how they see and interpret
  behavior.
• Disagreement can occur if behavior categories are
  not clearly defined.
• Strategy Establish interrater reliability.

13
Reasons to check interrater reliability

• 1. To check that observers are accurate and
  procedures easily reproduced.
• 2. Your observers meet some standard you
  established
• 3. Problems that you detect can be corrected
  with more training.

14
Simplest way to assess interrater reliability
is evaluate agreement.

• Count of times observers agreed.
• Divide this by total number of observations.
• Total of agreements x 100.
• Total of observations
• A percent around 70 is acceptable.
• Exact agreement or looser definition of
  agreement?

15

• Strategy
• You can average scores across observers
• Or one observer is named the main observer and
  the other is secondary observer. The secondary
  one is only to establish reliability.

16
(No Transcript)
17
(No Transcript)
18
Naturalistic

• Observe participants in natural environment
  without controlling or manipulating variables.
• Avoid changing situation that might affect
  natural, ongoing behaviors of participants.
• Properly conducted naturalistic observation has
  high external validity

19

• Even act of observing may disturb behavior of
  those you are watching.
• If disturb, then you decrease external validity.
• Habituate your participants to your presence
• Let your participants get used to you or camera
  before you begin observation.

20
Reactivity

• Individuals change their usual behavior because
  they know theyre being observed.
• On their best behavior
• Yet researcher wants to see normal behavior
• Strategy
• Conceal or disguise observer (people arent
  aware they are being observed -- must consider
  ethics)
• If observer is around long enough, participants
  get used to his presence
• Use indirect observation

21
Demand characteristics

• Participants try to guess what is expected, and
  behave like they think the researcher wants them
  to behave.
• Look for cues and information from researcher.

22
Ethnography

• Looks at cultural aspects and context of behavior
• Researcher becomes immersed in behavioral or
  social system being studied.
• Participant observation
• Functioning member of group
• Non-participant observation
• Observe as nonmember.
• Anthropology and sociology

23

• Members who know they are being observed may
  behave differently than they otherwise would,
  threatening external validity.
• Your interaction as member could create even
  more differences.
• Training to not interfere with the natural group
  process or using observers who are blind to the
  purposes of the study.

24
Gain access to group or organization you wish to
study

• (Shopping malls, vs. country club or police
  stations or classrooms (need permission)
• Try to get information about social structure,
  routines...
• Get info in library first
• Gain entry to group
• You might have to bargain with gatekeepers

25

• Make observations and record data
• Write notes, jot down thoughts,
• voice recorder,
• or wait till end and when alone write all your
  notes.

26

• Analyze data
• Look for themes, topics, issues
• Triangulate data
• from different sources, methods,

27

• Physical traces
• Cans in recycling bin
• University of Arizona study of dorm garbage
• Archival records
• Records of sports teams, stock market
• Govt statistics, census
• Public records - marriage applications
• Mass communications
• Analysis of Ann Landers/ Dear Abby letters
• Analysis of magazines targeted to teenage girls

28
Sociometry

• Identify and measure interpersonal relationships
  with a group.
• e.g. friendship patterns of 3rd grade students.
  Children identify who are their friends and who
  are not their friends.
• E.g. Identify if smokers associate primarily
  with smokers?