Observation

1
Observation
2
Single-subject design experience

• What tried to change?
• What interventions used?
• Successful?

3
Thinking about measurement

• Part of what you did involved measurement
• What was measured?
• How did you measure it?
• Was this a good system of measurement?
• How to determine?

4
Where psychology could get stuck

• Whats the funniest show on tv?
• Whats the most well-made movie of any time?
• Whats the best book ever written?
• People dont always agree
• Why not?
• Why could this be a problem?

5
The options

• Spend time debating what it means for a show to
  be funny
• Come up with an operational definition a means
  of turning a global, general concept into
  concrete, measurable terms
• May involve creating a scale score
• May involve making observations

6
What can be measured via observation?

• What variables can we measure by simply observing
  people?
• Need to be sure that operationalization of
  variable is specific enough, so that any person
  would make the same observation
• Need to develop a coding system a set of rules
  of what behavior gets translated to what measure

7
Checking the system

• Inter-rater reliability Do two different raters
  end up with the same measurement, if they use the
  same system and observe the same event?

8
People

• How much would you want to be friends with?
• Coding system?
• How smart?
• Coding system?
• How much would you trust your kids with?
• Coding system?

9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
Did the coding system help?

• What was good about it?
• What was bad about it?
• How could it be improved?

15
Quantifying things

• Inter-rater reliability
• In itself is a hypothesis
• Nullno association between coders ratings
• Variables who coded, what coded
• Can be quantified
• Depends on what type of variable measuring
• agreement
• correlation

16
Keys for inter-rater reliability

• Independent observations
• Observations of the same thing

17
Applying it all to McNulty Karney

• What variable was measured using observation?
• Why did they use observation?
• How did they test for inter-rater reliability?

18
Watching tv for homework

• Develop one variable to measure concerning tv
  (note one variable, not one hypothesis)
• Specify whether qualitative or quantitative
• Specify type of measure will use
• Specify requirements for each code
• Determine how to make sure measuring same thing
  (same show, same ad)
• Make sure have multiple observations (e.g., dont
  just each code one show, for overall aspects of
  show)
• More details on website will calculate
  inter-rater reliability and write paper on
  experience