ch 10 measurement and data collection Flashcards
5 concepts of measurement
-directness of measurement
-measurement error
-level of measurement
-reliability
-validity
operating principle: determining the value of
concrete things (O2 sat, temp, weight)
direct measurement
operating principle: the thing to measure is
abstract, a characteristic, or a concept (pain, depression, coping, self care, self esteem)
indirect measurement
how to calculate measurement error
observed score (in study) - true score (in real life) = measurement error
2 types of error
systematic error
random error
type of error: the variation in measurement is in the same direction
systematic error
type of error: the difference is without pattern
random error
4 levels of measurement from lowest to highest
-nominal (lowest)
-ordinal
-interval
-ratio (highest)
level of measurement: discrete category (diagnoses, gender, ethnicity)
nominal
level of measurement: ranking in order, numbers preserve order (levels of mobility, income, army ranking, levels of burns)
ordinal
level of measurement: equal numerical distances, can go below 0, represents a continuum (temperature, agree/disagree scale)
interval
level of measurement: includes all criteria and has absolute zero point, continuum, exclusive and exhaustive (weight, BP, blood glucose)
ratio
types of validity
-construct
-content
type of validity: how well the concepts and
definitions match each other
construct
type of validity: does it truly represent the
concept under study
content
How consistently the measurement technique measures the concept of interest.
-exists in degrees
-expressed by coefficient (r)
reliability
3 aspects focused on in reliability testing
-stability
-equivalence
-homogeneity
Lowest acceptable coefficient for well-developed measurement tool
0.80
acceptable coefficient for newly developed instrument
0.70
Statistical representation of degree of relationship between two sets of measurements
correlation coefficient
The accidental discovery of
something useful or valuable;
can lead to new insights and must be named in study
serendipity