Week 5 Data collection & Level of measurement Flashcards
What is Data collection?
Process of acquiring subjects
Collecting Data
Must be clearly report
3 Key considerations of data collection
- Sample size
- Control over extraneous factors
- Consistency in collecting data
Instruments/Methods to collect data
- Self reports
- Observations
- Biophysiologic measures
Self reports
Interviews, questionnaires, scale
Disadvantage of self reports
Prone to response set biases
Examples of response set biases
- Social desirability
- Extreme response
- Acquiescence response
Social desirability bias
Tendency to misrepresent attitudes/traits by giving responses that are consistent with prevailing social views
Extreme response bias
Tendency to consistently express extreme attitudes/feelings no middle ground
Acquiescence response bias
Tendency to agree with all questions regardless of content
What are structured & unstructured observations?
Structured - follow checklist
Unstructured - happens spontaneously
Biophysiological measures
Most objective
e.g BP, temperature, blood level
What to consider when doing observations?
Do inter rater reliability - to be consistent in findings with other observers
What is measurement?
Process of assigning numbers/values to individual’s health status, objects or situations using a set of rules
What are the 4 levels of measurement?
- Ratio
- Interval
- Ordinal
- Nominal
What are the 2 levels of variable?
Numerical (quantitative)
Categorical (qualitative)
What is a nominal scale & its example?
Simply assigning a category name where numbers are arbitrary & have no real meaning
e.g gender (1=male, 2=female), diagnosis
Rules of a nominal scale
- Cannot be rank ordered
- Must be exclusive & exhaustive categories
What is an ordinal scale & its example?
The numbers used to identify the attribute have meaning
e.g what is the highest degree of level of education
Rule of ordinal scale
Categories can be ranked
Order of the numbers matter but with unequal intervals
What to analyse in ordinal scale?
Median - select number that is exactly in the middle where numbers are in value order
What to analyse in nominal scale?
Mode & frequency
What is an interval scale & its example?
Equal numerical distances between intervals, where numerical value means something on the real number scale
e.g temperature, likert scale scores, time
Rules of interval scale
Order of numbers matter
Lacks a zero point or fixed beginning (no true zero)
What to analyse in interval scale?
Mean - (addition/subtraction) average of all respondent’s answers
What is a ratio scale & its example?
Highest form of measurement
Continuum of values with an absolute/true zero point
e.g weight, money
What to analyse in ratio scale?
Mean - addition/subtraction/multiplication/division
Advantages of measurement
- Removes guesswork in gathering & communicating information
- Obtains more precise information
- Provides a language for communication & analysis
What are the errors of measurement?
- Situational contaminants
- Response set biases
- Transitory personal factors
