test 1

Question 1

what is multiface?

Answer 1

many phenomena are not simple. Especially when it comes to people, could be a lot of errors.

Question 2

The importance of studying measurements?

Answer 2

• minimizing subjective judgment: subjective judgment can cause errors, critically evaluate
• make more precise statements : common language, keep it simple, no misunderstanding
• quantify your observations: to what extent is this real

Question 3

who is francis galton?

Answer 3

the father of psychometrics, obsessed with measurements.
calculating the degree of association between any two characteristic.
Individual difference: degree of characteristic understanding the ways in which people differ
Karl Pearson: his student

Question 4

traits vs states

Answer 4

traits: extraversion, IQ, depression, anxiety
resistant to change over time
refer to behaviour in general
often easier to measure with QUESTIONNAIRES

State: sleepiness, hunger, depression, anxiety
Subject to change over time
refers to behaviour in the moment
easy to measure with TASK AND QUESTIONNAIRE

Question 5

Why study psychometrics?

Answer 5

Ensure reliable & valid measures: Reliable; consistency, take a test multiple times & always get the same results.
Validity: a measurement should measure what it is suppose to measure.
Application: good judgments require good measurement
questionnaire are good: make good dependent variables. Eliminate error as covariates controls, experimental groups.

Question 6

Scaling methods (3) 
3 basic approaches to scaling

Answer 6

1. Stimulus-centered: relates physical, directly measurable, stimuli to perception…
   Individual perception, basic perception. ex, how bright is the light.
   Convert a physical known quantity of sum phenomena to a psychological representation perception base. DOES NOT APPLY TO QUESTIONNAIRE, WILL NOT FOCUS ON THIS.
2. Response-centered: relates individuals along a psychological continuum. includes ranking the order of a series of choices, assessments of similarities of objects, etc. (harder to work with)
3. Subject entered: estimate the subjective presence, absence or degree of a construct.
   Getting a numerical score to identify the amount/quantity of a construct the individual possesses.
   Trying to learn something about the thing i am interested in & putting in down on a degree of the construct. Come up with a numerical score. more likely to get a number of the degree the person reflects.
   .

Question 7

what is a construct?

Answer 7

things you are interested in measuring

Question 8

four types of measurements

Answer 8

Nominal
Ordinal
Interval
Ratio

Question 9

Nominal

Answer 9

The number has no value, it is assigned as a label.

Question 10

Ordinal

Answer 10

Not just labels, it ranks individuals. We must use numbers in a meaningful way. ex. Order of the smallest to the tallest.
It indicates an individuals or objects value based on its relationship to others in the group.
IMPORTANT TO REMEMBER:
1) Rank has only meaning within the group
2) gives no information about how closely two individuals or objects are related.
NO MEAN OR STANDARD DEVIATION

Question 11

Interval

Answer 11

Numbers are labels, reflect rank and tell us exactly how much more of something we have now.
Each point of the likert-typa rating scale represents an equal distance or amount of the construct being measured.
-No point at which attribute is totally absent cannot be specified.
MEAN AND STANDARD DEVIATION CAN BE CALCULATED

Question 12

Ratio

Answer 12

numbers are labels, reflect rank & tell us exactly how much more of something we have and a score of zero is the smallest possible number.
Ratio scaled numbers are essentially the same as interval scaled numbers except they cant be negative.
They also have a point that represents an absolute absence of the property being measured and that is zero.

Question 13

Standardize correlation

Answer 13

standardize measurements about converting our numbers into this common z distribution of score. A small negative (ex. -4) would indicate a very low score on the distribution where 4 would be a very high score of the distribution. Standardize to make a common reference point.

Question 14

Four basic steps to z-score and standardization

Answer 14

1) Provide a new measure to a very large sample.
2) verify data represent full range of scores
3) determine mean and standard deviation
4) calculate z scores (number of SD)

Question 15

A critical part of the research process

Answer 15

• Deciding which characteristics must be measured to assess your construct
• Examining existing scale, inventories, test, questionnaires for potential items/methods.
• Verifying the utility of your items/methods

Question 16

4 steps to make a research questionnaire

Answer 16

1) Choose a construct : what is the thing your going to measure.
2) Make items: review existing literature, questionnaires/tests, consult with experts, make some questions
3) Response format: what response format do you want to use? Different formats for some questions?
4) Pilot test: find out which items work best, reduce items and repeat, few times.

Question 17

Making items

Answer 17

• how you word your questions, and potential consequences
• deciding how many questions you need for each aspect of your construct
• deciding whether to reverse-score items
• decide to include redundancy

Question 18

Redundancy

Answer 18

strengthen a scale, rephrasing of items in similar ways brings out their common cause

Question 19

Name all the types of scales ( 6)

Answer 19

true/false
multiple choice
thurstone scale 
guttman scale
analog scale
likert scale
likert-type scale

Question 20

Thurstone scale

Answer 20

statements are carefully calibrated at specific levels of construct. Potentially equal intervals.
Only agree with one & disagree with the other three.
Score is the one you agree with.

Question 21

Guttman scale

Answer 21

improved version of thurstone scale. Agree with two and disagree with the other two. Can be reversed. Your score is the numeric value of the transition point.

Question 22

Analog scale

Answer 22

similar to a semantic differential scaling formate, except you can only choose specific value along the continuum, the analog can be anywhere.

Question 23

Likert scale

Answer 23

strongly disagree to strongly agree, 5 choices but you need a neutral middle scale.

Question 24

Reverse scoring

Answer 24

when we need/want to ask questions that reflect the opposite of our construct.

• Reverse the number scale
• Can be useful to break participants habits
• Problematic when it is not easy to define

Question 25

writing good questions

Answer 25

• don’t use double negatives (bad for construct validity)
• avoid using double-barreled questions.
• avoid leading questions

Question 26

types of response biases

Answer 26

Acquiescence: yeah-saying
social disability
malingering: answering in a way that makes you look bad
extremity

Question 27

What happens with poor measurements&

Answer 27

• not valid measure
• limits the ability to assess the degree of association between the two things.
• may not cover a full construct & prevent accurate conclusion