W11: Measurement and Sampling

Question 1

What is measurement?

Answer 1

fundamental part of scientific research; assignment of values (numbers) to objects, events, outcomes

process of observing, recording observations that are collected
requires rules, isomorphism, indicators of control, operational definition/procedure for measuring

Question 2

Give an example of measurement for each type: objects, events, outcomes.

Answer 2

objects: hospitals (per geograph. radius)
events: deaths
outcomes: quality of life

Question 3

What are the different levels/scales of measurement?

Answer 3

-nominal
-ordinal
-interval
-ratio

Question 4

Define nominal level

Answer 4

numbers are used as labels where they have an inherent numerical value eg. gender, ethnicity

Question 5

Define ordinal level

Answer 5

numbers are a rank order and are assigned a rank value

Question 6

define interval level

Answer 6

10-20, 30-40, etc.
knowing rank order and exact distance between each observation

Question 7

define ratio level

Answer 7

type of number that has true zero

Question 8

Why is the level important?

Answer 8

helps you decide how to interpret data from variable
helps you decide what stat analysis is appropriate on values that were assigned

Question 9

What is reliability?

Answer 9

occurs when measurement tool measures same thing more than once and results in same outcomes
aka consistent

Question 10

What are the types of reliability?

Answer 10

test- retest reliability
parallel forms reliability
inter-rater reliability

Question 11

Define test-retest reliability. What is the method?

Answer 11

measures how stable a test is over time (keep conditions similar for both testing times)
method: do same test at 2 different times to same group of participants
correlate scores at time 1 with scores at time 2

Question 12

What is parallel forms reliability? What is the method?

Answer 12

measure of how equivalent 2 different forms of a test is
method: administer 2 different forms of same test to same group of participants
- correlate 2 sets of scores

Question 13

what is inter-rater reliability? What is the method?

Answer 13

measure of consistency (agreement) from rater to rater
method: have more than 1 rater rate the same thing and correlate the scored between them

Question 14

How can you increase reliability?

Answer 14

-eliminate items in a test that are unclear, ambiguous, faulty, vague
-standardize conditions under which test is taken
-minimize effects of external events so that true test performance is not affected
- maintain consistent scoring procedures
-standardize instructions to respondents

Question 15

What is validity?

Answer 15

occurs when test/instrument measures what it is intended to measure; refers to results of a test
aka truthfulness

Question 16

What is external validity?

Answer 16

research results apply to people and situations beyond particular sample studied

Question 17

What is internal validity?

Answer 17

degree to which an experiment is designed so that a causal relationship can be demonstrated

Question 18

What are the different types of validity?

Answer 18

-face
-content
-criterion
-construct

Question 19

What is face validity?

Answer 19

declaration by researchers claiming instrument is reasonable successful in measuring what it is intended to measure
- based on expert opinion

Question 20

give an example of face validity.

Answer 20

questionnaire that is supposed to assess seniors’ nutritional status, you may recruit a few experts in seniors nutrition to review it and judge whether they think items are meaningful
- items about accessibility to food>seniors’ income levels (not direct enough)

Question 21

What is content validity?

Answer 21

measure of how well items in measurement tool represent entire universe of relevant items
based on expert opinion and theoretical perspectives

Question 22

Give an example of content validity

Answer 22

measuring QOL in seniors, need comprehensive range of Qs- ensure at least one Q is about appetite because seniors view appetite as key part of QOL.

Question 23

What is criterion validity?

Answer 23

measure of extent to which test is related to some criterion
-requires prospective study design; aka predictive validity

Question 24

Give an example of criterion validity

Answer 24

breast self examination (“test”) for predicting a breast cancer diagnosis at some later time

Question 25

What is construct validity?

Answer 25

measure of how well measurement tool items are tapping into underlying theory, construct, or model of behavior
- tested over time and conditions to see f it hold true in all instances

Question 26

Give an example of construct validity

Answer 26

does an intelligence test actually measure intelligence?
-no gold standard
- test can discriminate