NURS 2031 FINAL EXAMACTUAL QUESTIONS AND

NURS 2031 FINAL EXAM(ACTUAL QUESTIONS AND

VERIFIED ANSWERS LATEST GRADE A+

(step 3 in scale development):

What problems should you be watching out for in the language used in the scales? - ---Answers-----jargon

-double-barreled questions: asking more than one question in

one

-vague quantifiers: e.g. often, seldom, rarely etc.

(step 3 in scale development):

What is cognitive interviewing/how does it help develop scales? - ---Answers-----it involves asking people to read the terms and Answers them into their own words OR asking people in interview format to answer the questions and explain their Answers.-this can help identify problems in the questions such as difficulty of the Qs, comprehension, clarity, misinterpretation etc.) -normally done with 10-15 people.

(step 3 in scale development):

What are some biases that can be present when responding? - ---Answers-----social desirability bias -yea-saying/acquiescence bias (responders agree with any statement even if they don't actually) -end-aversion bias (avoids extremes of the scale) 1 / 4

-cultural factor (some people from certain cultures may be more reserved/weary of offending - yea-saying/some may be more expressive and likely to choose extreme ends)

(step 3 in scale development):

What are some ways to scale the items (display the options) -

---Answers-----categorical judgements:

-continuous judgements

What is a categorical judgement scale? - ---Answers----- choosing from a number of options/alternatives -e.g. selector religion - catholic/christian/jewish/buddhist

What is a continuous judgement scale? - ---Answers----- indicates response along a continuum

-2 kinds:

1) unipolar - starts from 0 and goes only to 1 extreme (no natural midpoint) 2) bipolar - extremes at both ends and has a natural midpoint (ideally 5-9 categories/should have an odd number if there is a middle point)

What are some pros and cons of bipolar continuous judgement

scales and examples? - ---Answers----Pros:

-can be used in children (8+)

Cons: 2 / 4

-not all boxes are labelled and the labelled boxes are often picked more than non-labeled) -there are biases against picking negative boxes -earlier responses often influence later responses

Examples:

-visual analogue scale: e.g. pain to no pain on a line.

-adjectival scale: e.g. what should the family's role be in life

support? - boxes; none/minor/major/soledeciders -likert scale

-harter scale: not as much cognitive demand/2 options - e.g.

some kids have friends: boxes: true for me/not true for me.

(step 4 in scale development - selecting the items):

why do you test the scale first with >50 people (10 people per item on the scale ideally) / what do you do with this information? - ---Answers-----to ensure that all responses are used by some.-you should eliminate the items which were not used -eliminate items which more than 90% gave the same response.

(step 4 in scale development - selecting the items):

What is a factor analysis and how is it used to analyze the scales. - ---Answers-----advanced statistical procedure used to determine if a large number of variables (such as items on a scale) can be grouped into a number of subgroups in which the 3 / 4

variables correlate highly with each other and poorly with other items.

-Factor analyze the scale:

1) because item grouping makes clinical sense 2) loading (correlation) of each item is at least 0.35 (moderate) to consider keeping.3) item should load (correlate) on only one factor.

-if any are not met - scrape and rewrite the item.

What is random error vs. systemic error? - ---Answers-----

random: unpredictable error

-systematic: bias that affects all scores in one direction

Explain the ways that reliability may be determined:

-test-retest reliability -inter-rater reliability

-internal consistency - ---Answers-----test-retest: (stability)

do we get the same results over time

-inter-rater: (equivalence) would two or more people giving

rating get the same results

-internal consistency: degree to which item in the scale are

correlated/to do with the items/do the items belong together.

- Cronbach's alpha: measurement of the degree of correlation

= the more items the better/the higher coefficient the better.

• / 4