Quiz 1 Flashcards
Reliability
Is the extent to which an experiment,test, or any measurements procedure yields the same result over & over
Types of reliability: (3)
Inter-rate reliability : the degree of agreement between 2 or more
Test-retest: how much does your score stay the same when taken twice?
Internal consistency: how much do item on a surgery measure the same thing
Validity
Is the extent to which a measurement instrument measure what is it suppose to ( how accurate )
Types of validity
Face: do the items in the measurement instrument looks like they are measuring the construct
Content: to which the measurement instruments cover the entire range of meaning within a concepts
Types of validity
Criterion related: degree to which a measurement intrusment relates to another measure of the same variable or a future outcome
Construct: the degree to which a measure relates to other variables characterizing the same theoretical construct & is different from unrelated ones
What makes a good question?
- open & closed questions
- clear, simple
- culturally sensitive
- avoided doubled barreled questions
- avoid double negatives
Questionnaire instructions ( quantitative )
- include clear instructions & introductory comments when appropriate
- spread out /uncluttered
- contingency questions
- readable language
- questionnaire length
What is a scale ?
Variable may be too complex to be measured only one item, responses to individual items are summed to a scale score (1-5)
Qualitative instruments
Informal, conversational interviews with unplanned, no pre determined list, flexible
Use neutral probes (silence /repeating words)
-
Quantitative instrument
Objectivity, consistency, generalizability, test hypothese
Survey modes
- self administered: mail, electronic
- interviews: face to face, phone
Sampling frame
A list from which you can select your sample
Probability sampling
Random sampling
- avoid biases
- more representative then not probability
- enables estimate to made of accuracy
Systematic sampling ( type of probability )
Every third
Stratified random (Probability)
Strata - homogenous mutually exclusive groups within population
Groups you can create based on distinguishing characteristics of your population
Put into different boxes (age sex) before
Cluster sampling
P
Naturally occurring groups ( high school classrooms & neighborhood )
Multi stage cluster
P
Going through multiple steps to get your sample
Geography
Convenience sample
NP
Convenient & readily avaible subject are included in the sample
Snowball sampling (NP)
Existing study subject recurit their friends
Judgement sampling (NP)
Chosen the sample base on who they think would be best
Quota sampling (NP)
Divide sample into exclusive groups
Select from these groups using purposive non probability sampling
Types of qualitative interviewing
Open ended interview ,informal conversational , & interview guide
Informal conversational interviewing (qual)
Is an unplanned & unanticipated interaction between the two tha occur naturally
Standard open ended interview (qual)
When resources limitations leave insufficient time to purses less structures strategies