Measurement & Scales - Data Types

Question

What is a characteristics of nominal data?

Answer 1

- it can be grouped into two or more categories that are mutually exclusive (something can not fit into more categories, it has to fit in ONE) and collectively exhaustive (it has to fit in a category)

- least powerful, because they only suggest classification, they suggest no order, distance relationship and have no origin.

- nominal scale waste information about varying degrees of the property being measured.

- you count the number of cases (e.g. # men/female, #category a/b/c/) You can conclude which category has most members, that is all.

- there is no measure of dispersion for nominal scales.

Answer 2

- it is sufficient if you are not interested in precise measurements but mroe in explorative work/relationships

- cross-tabulations mgith provide insight to patterns

- used a lot in post-facto reseach

Answer 3

- power of nominal (classification) and ordinal (order) data + one addition.... they incorporate the concepts of equality of interal! (distance between 1 and 2 equals distance between 2 and 3).

However, there is no origin.

time between 3 and 6 is same amount as time between 6 and 9. But 6 is not twice as late as 3 because there is no 'zero time'.
Same with temprerature.

You use aritmic mean as measure of central tendency. Standard deviation = measure of dispersion.

Also t-tests, F-tests and other tests can be applied.

Answer 4

Conversint or recaling a variable involves reducing th emeasure from the more powerful and robust level to a lesser one. E.g. comparning dichotomous nominal variable (male female) with ratio-data salary, you can rescale ratio numbers to high-low.

Answer 5

- participant error: dimensions should be measured (e.g. employee status, gender, social class, etc), also a participant might just 'guess'. also their mood might have impact.

- situational error: if another person is present, if they fear anonimity,

- measurer-error: stereotypes in appearance, prompting smiles, or wrongly processing (checking the wrong box)

-data-collection instrument error: mechanical effects (inadequate space for replies), difficult words, or poor selection from universe of content items (it might not collect all data).