Lecture 2: Measuring Behavior

Question 1

Self-Report Measures

Answer 1

operationalize answers by using peoples answers to questions about themselves

Question 2

What are observational Measures?

Answer 2

records observable behaviors or physical traces of behaviors (ex: happiness based on smiling by recording the number of times a person smiles)

Question 3

What are physiological measures?

Answer 3

Record biological data
ex: brain activity, heat rate, special equipment is needed such as EEG or MRI machines

Question 4

Researchers often use ________ of measures

Answer 4

triangulation

which is the use of multiple approaches to measure the variable of interest

this is because they hope to capture the construct of interest as fully and best as possible

Question 5

Describe the nominal scale of measurement

Answer 5

The nominal scale of measurement is a measurement used in research that is the process of assigning numbers to participants responses in order to perform future data analysis.

This specific scale is where numbers assigned act as labels. There is no significance between the labels nor is their ranking value to the labels.

ex: gender, ethnicity, race

Question 6

Describe the ordinal scale of measurement

Answer 6

The ordinal scale of measurement is where numbers assigned to participants responses/ scores are ranked in some order.

One score signifies higher value/ more strength than another

ex: 10 is higher than 7

ex2: the order can tell who finishes a test first however cannot tell how much faster participant one finished before participant 2

• distance between scores is insignificant/ the same

Question 7

Describe the interval scale of measurement

Answer 7

interval scale of measurement has equal distance between numbers, and there is no true zero meaning that a score of 0 cannot be obtained (the absence of something is not possible; ex: temperature)

ex: iq tests scores A: 30 B:60 C: 90 the distance between each of these values is the same and can be compared

Question 8

Describe the ratio scale of measurement

Answer 8

Ratio scale measurement is the highest measurement that can be obtained

comparisons between scores can be made such that it can be stated that student A scored 2x as high as student B

In ratio scale there is equal distance between the numbers and there is a true zero

ex: years of work experience, number of children in a household

ex2: number of exam items answered correctly

Question 9

What does the type of measurement scale determine?

Answer 9

the amount of information provided by the given measure and the statistical analyses that can be run

Question 10

What is reliability and how is it assessed?

Answer 10

reliability refers to how consistent the results of a measure are

can they be replicated?

assessed via correlational coefficients

Question 11

What are the three ways reliability can be assessed?

Answer 11

Test retest: when retesting roughly the same score is obtained

*most relevant when measuring constructs we assume to be stable (ex: weight at the beginning and end of class)

interitem: internal reliability; a person provides a consistent set of responses no matter what the phrasing of the item is (consistency is looked at on a scale ex: diners happiness scale- items themselves hang TOGETHER)

interrater: two or more independent observers come up with consisten findings

ex: the number of times children cry while playing at a playground is observed by you to be 30 and by your co-researcher also ~30

Question 12

Describe a correlation coefficient

Answer 12

a correlation coefficient is a measure of the extent to which to factors vary together

represented by r

Correlation coefficients sign and magnitude are taken into account

• positive correlation coefficient indicates that the factors vary together in tandum ; same direction
• negative correlation coefficient indicates that there is an inverse relationship between factors/ variables; as one increases the other decreases; opposite direction

The larger the CC the stronger the relation between variables (+1 to -1)

0= no relation

Question 13

Cronbachs alpha measures what?

Answer 13

interitem reliability

the closer the value is to 1 the better the scales reliability

ideally cronbachs alpha will be above .70

an item that is not above a 0.70 will be dropped

Question 14

What is validity ?

Answer 14

refers to whether a measurement captures what it is intended to measure

a measure might be valid for one purpose but not for another

Question 15

Describe the three types of validity

Answer 15

face validity: a measure that appears to measure what it attends to measure
*based on subjective judgement from the researcher

construct validity: How well a conceptual variable is operationalized
*other relevant measures need to identified

*opposite is discriminant validity- don’t want the measure to be correlated with unrelated measures

criterion-related validity: whether the measure is associated with a concrete behavioral outcome
*a relevant behavioral outcome needs to identified

two subtypes: concurent: measures behavior at the time simultaneous to administration

predictive: measures whether it can distinguish between people over time ex: SAT scores and success in college

Question 16

Test Bias occurs when?

Answer 16

a measure is not equally valid for all participants/ test takers

test bias is often very subtle

bias tests have stronger predictive validity for some groups than for others

some factors that result in test bias are culture and unclear questions

Question 17

what are ways to minimize test bias?

Answer 17

informants review the items and provide their opinions on them

read through the questions and inform if they understand the question

Question 18

reliability is what?

Answer 18

consistency or dependability of a measuring technique

Question 19

how do you know if a measure is considered sufficiently reliable for research purposes?

Answer 19

cronbachs alpha is used
adequate interitem reliability if cronbachs alpha is over 0.70

Question 20

describe test retest reliability

Answer 20

consistency of participants responses on a measure over time

it is determined by taking two measurements separated by x amount of time

only works on measures that are not expected to change in such short amount of time

that being said hunger is not a great thing to test

Question 21

what is validity?

Answer 21

the degree to which variability in participants scores on a particular measure reflects variability in the characteristic we want to measure

• do scores on the measure relate to the behavior or attribute of interest?
• are we measuring what we think we are measuring?

Question 22

what is construct validity?

Answer 22

the degree to which the construct of a measure correlates as expected to constructs of other measures

Question 23

what is convergent validity?

Answer 23

a measure correlates with other measures of related constructs

Question 24

what is discriminant validity?

Answer 24

the measure doesn’t correlate with measures of constructs that its not related to