Textbook Notes (280,000)

CA (170,000)

Brock U (1,000)

PSYC (200)

PSYC 2P25 (4)

Michael Ashton (4)

Chapter 1

This

**preview**shows page 1. to view the full**5 pages of the document.**Chapter 1: Basic Concepts in Psychological Measurement

Some Simple Statistical Ideas

Levels of Measurement

There is not usually a meaningful “zero” level of a psychological trait

In psychology, we are not able to describe ratios between people’s levels of a variable or a

person’s absolute amount of a variable

Ways of measuring

o Ranking people

o Observation

Standard Scores

Measure the differences between scores

o Differences in numbers used for measuring variables might cause difficulties when we

want to compare someone’s scores across two or more traits

o So we need some way to relate scores on one scale to scores on another scale, so that

we can compare levels of one characteristics with levels of another, or to compare

scores on the same characteristic as measured by different scales

Scores are converted into standard scores

o The first step in calculating a standard score is to take an individual’s score on a given

scale, and then subtract the mean score for the person’s that have been measured. This

difference tells us whether the person is above average or below average.

o The second step is to divide this difference (between the person’s score and overall

average) by the standard deviation, a number that indicates how much variability there

is among a variable

o The result of the two steps is to give a universal or standard way of expressing people’s

scores on a given characteristic, regardless of the original distribution of scores on that

characteristic

Standard scores have two special properties

o The average score on a standard-score scale is exactly one. That way after we have

calculated the scores, we can meaningfully compare a person’s scores across different

variables

Correlation Coefficients

After we have calculated the scores, we can easily figure out the extent to which the two

variables “go together,” or correlate

Correlation coefficient = r

o It can have values ranging anywhere from a maximum of +1 to a minimum of -1

o +1 perfect positive; -1 perfect negative

###### You're Reading a Preview

Unlock to view full version

Only page 1 are available for preview. Some parts have been intentionally blurred.

o Between +.20 and -.20 is a weak correlation

o Between +.20 and +.50/-.20 and -.50 is a moderate correlation

o Above +.50 or below -.50 is an extreme correlation

The binomial effect size display (BESD) is a table that helps to give us an intuitive understanding

of the meaning of a correlation of a give size

o Contains two rows and two columns

o One row represents the set of people who have a “high” or above-average level of one

variable, and the other row represents the “low” or below-average

o One column represents the set of people who have a “high” or above-average level of a

second variable, and the other column represents the “low” or below-average of that

second variable

Assessing Quality of Measurement: Reliability and Validity

Reliability

The extent to which it agrees with other measurements of the same variable. When there is a

good agreement between measurements, this tells us that they are assessing some real

characteristic, rather than meaningless numbers.

There are several ways that reliability can be assessed

Internal-Consistency Reliability

o When evaluating the quality of a psychological measurement, we need to consider the

error that results from differences among the “items” or parts of the measurement,

such as the various questions on a test or scale

o To the extent that an item measures some specific variable of its own, rather than the

characteristic that we are trying to assess, we say that the item has “error” variance. If a

test or scale overall has a large proportion of error variance, then it cannot be

measuring any single, common characteristic reliably.

o The reliability of a score that is found by averaging responses to several items basically

depends on two things:

The number of items

If we are averaging out people’s responses to items that have

something in common, that common element will become stronger and

strong when we add more and more items, so by averaging out across a

larger number of items, we get a more reliable score.

The correlations between items

If we are averaging out people’s responses to items that have

something in common, that common element will be stronger to the

extent that the items are correlated with each other, because those

correlations tell us how much each individual item is measuring the

common characteristic.

###### You're Reading a Preview

Unlock to view full version