Psychology : Psychometric Report

profileKyra2795
PSY4046andPSY3018Week17-ReadingWeek-NormalisingEQscores.pdf

PSY4046 AND PSY3018

READING WEEK (WEEK 17)

TAKING THE EQ TEST

Dr Jackie Meredith

1

MEASURING EMOTIONAL

INTELLIGENCE (EQ)

2

Think about the questions on each of the next three slides – this relates to

normalising scores.

Comparative published measures Measure Correspondin

g Theorist

Mode of Measure Brief Description

Mayer-Salovey-Caruso

Emotional Intelligence

Test (MSCEIT)

Mayer, Salovey Performance

based

Specific tasks to measure level of ability in

each of 4 branches of emotional intelligence

Emotional Quotient

Inventory (EQ-i)

Bar-On Self-Report 133 self report items measuring total EIQ

and 5 components

Emotional Competency

Inventory (ECI)

Goleman Self Report &

Other Report

Multi-rater instrument that provides ratings

on a series of behavioural indicators of

emotional intelligence

Emotional Intelligence

Appraisal (EIA)

Goleman Self-Report &

Other Report

7-minute assessment meant to measure the

existence of Goleman’s 4 components of

emotional intelligence

Work Profile

Questionnaire-Emotional

Intelligence Version

(WPQei)

Goleman Self-Report 84 item measure of 7 of Goleman’s

competencies though of as most essential

for effective work performance

Levels of Emotional

Awareness Scale (LEAS)

Other Self-Report Measures levels of awareness of emotions

in oneself and others

Self-Report Emotional

Intelligence Test (SREIT)

Salovey &

Mayer or Other

Self-Report 33 item measure of Salovey and Mayer’s

original concept of emotional intelligence

3

EQ TESTS 1: ECI-360 - GOLEMAN

• Normed using 6000 respondents in America and UK

• Mainly white males holding mid to senior-level management positions

• Scores on the ECI significantly positively correlated with:

• salary (r = .40)

• job success (r = .33)

• life success (r = .46; • Sevinc, 2001

4

Q: How generalisable is this norm population?

Q: What is the target group for this test?

Q: Which type of validity does this test demonstrate?

EQ TESTS 2: MSCEIT

• The Mayer-Salovey-Caruso Emotional Intelligence Test

• Normed on a sample of 5,000 men and women • 50 research sites worldwide

• Mainly white women under 30…

• Good Predictive Validity • MSCEIT predicted social deviance in a sample of 207 college

students • involvement in physical fights and vandalism

• even after controlling for personality and IQ • Brackett & Mayer, 2003

5

Q: How generalisable is this norm population?

Q: What is the target group for this test?

Q: Which type of validity does this test demonstrate?

EQ TEST 3: EQ-I – Bar-On

• Normed using 4000 Americans and Canadians • Normative sample 79% white and below 30 in age (equal males and

females)

• Bar-On’s (2002) EQi accurately differentiated between: • Those who were successful and those who were unsuccessful in

business and industry settings.

• High or low self-perceived success in military school

• Air Force Recruiters who were the most successful in their work

• Academically successful and unsuccessful university students.

6

Q: How generalisable is this norm population?

Q: What is the target group for this test?

Q: Which type of validity does this test demonstrate?

THE WECQ - Meredith

• Currently scores are normed using 581 British psychology undergraduate students. • What are the features of this population?

• Measures 5 key areas of emotional competence

• Important points:

• 1. Emotional competency is not set in stone. • This is something you can work to change.

• 2. The WECQ is primarily a measure of workplace emotional competency.

• Many items are based around workplace issues • However there will be aspects which relate to other areas of life:

• relationships – friends and family

• To some extent this depends how integrated you are as a person. 7

IMPORTANCE OF NORM GROUPS

• Of course there have been further validations of these

three major EIQ tests

• Using different Norm groups

8

• PSY4046 - You will need to write in your report about the choice of norm

group for the WECQ, which your participant is taking.

• Describe the features of the population (apart from all being

undergraduate psychology students) – age, sex, ethnicity?

• Why is this appropriate for the client?

• PSY3018 - You will explain in your HR report the importance of relating a

candidate’s scores to an appropriate norm when calculating scores – what

could be the consequence of using an inappropriate norm group?

To Think about

• If EIQ is an integrated concept, what does this mean in terms

of scoring? Might scores in one area correlate with scores in

another?

• Is personality an integrated concept?

• Can you have a ‘total’ personality score?

• If not, why not?

• There has been some research that suggests women have

higher EIQ than men – why might that be and in what areas

might you expect women and men to score differently?

9

SCORING

• First you will need to DOWNLOAD the Excel spreadsheet to calculate your RAW SCORES

• There are 7 TABS: • 5 for Domains of Emotional Competence

• 1 for ‘Other Items’

• 1 for Total EQ.

• As before you need to put the NUMBER in the cell which relates to the column

10

In Excel, all scoring is automatic but you must type NUMBERS in the columns, like for Personality

Place the

number

value of the

column in the

appropriate

boxes on the

grid and the

totals will be

calculated for

you – for

example in

top column 1

insert 1

instead of x,

in the second

column the

value is 2, so

insert 2…

Other Items

• One of the tabs is to calculate your score for ‘Other Items’

• You don’t have to calculate a Normalised score for this -

• Raw scores are fine.

• Don’t worry about this score – we will be discussing this in class next week!

• Don’t forget to bring your scores (including your RAW SCORE FOR ‘OTHER ITEMS’ to the class.

12

Calculating WECQ norms

YOU WILL NEED TO APPLY THE SAME TWO FORMULAE YOU USED FOR PERSONALITY TO EACH RAW SCORE:

• Z score formula:

Z = (x – μ)

σ

T score (standardised score) formula:

T = (Z score x 10) + 50

13

Different means and SD of course…

5 Domains +

an overall

EIQ Score

But First… Demographics

• For example there is evidence from previous research that PERSONALITY scores may vary across cultures

• Reflecting the value society puts on different behaviours

• For example Individualist vs Collectivist cultures

• Cultural norms may therefore be valuable when scoring personality

• We don’t have cultural norms, but we do have a multi-cultural sample, which is the next best thing…

14

Past research should always be consulted when deciding what demographics to collect during a study

But First… Demographics

• What OTHER issues from previous research may impact on EQ?

• Expressing emotion (girls more expressive than boys in child studies)

• Empathy (girls more empathic than boys, women more empathic than men in some studies)

• But the WECQ is not measuring empathy…

• What other demographics might affect EQ scores?

• How would we know?

15

Look at contemporary research…

ISSUES WITH MEASURING EIQ

• Some previous research shows a positive correlation

between EIQ and Education level

• So as we use UG norms, raw scores are likely to be fairly high.

• As there is also some evidence of sex differences (females

scoring higher than males) – so should we use different

norms for males and females in our study?

• How can we decide?

16

ANSWER = Look at the evidence from norm group…

HOW? = By examining Distributions and doing independent groups t-tests with SEX as factor…

SELF-INSIGHT

17

No

significant

difference

between

male and

female

means

Communication

18

No

significant

difference

between

male and

female

means

Judgement

19

No significant

difference

between male

and female

means

Resilience

20

No significant

difference

between male

and female

means

FOUNDATIONS

21

Males

scoring

significantly

higher than

females,

p<0.05

WOW! We didn’t

expect this…

WECQ TOTAL EQ SCORE

22

No significant

difference

between male

and female

means

Comparing with Population Norms

• Calculating your Z score

Z = (x – μ)

σ

• Calculating your T score (standardised score)

T = (Z score x 10) + 50

• This is the accepted formula in education and is used to calculate IQ norms, SAT norms, etc. from Z scores

• Ravid, R. (2005). Practical Statistics for Educators (3rd ed.). NJ: University Press of America.

23

 4 out of 5 scores and total score shows no differentiation between groups

 PLUS emotional intelligence is an integrated concept

 DECISION: We will use total sample norms

 there no overall differentiation between males and females in British psychology undergraduates in our sample.

THE STATISTICS FOR EIQ

You need to apply the same formulae to the 5 Domains and the Total EIQ Score.

• WECQ Domains: • Self Insight …………….. Mean: 75.29 SD: 8.68

• Communication … …….. Mean: 76.75 SD: 8.55

• Judgement .........………Mean: 78.06 SD: 7.88

• Resilience ..….................. Mean: 67.24 SD: 10.02

• Foundations ...................Mean: 69.90 SD: 10.92

• TOTAL EIQ ............. Mean: 73.44 SD: 9.21

24

Note – smaller standard

deviations than for personality

– less variation

Note – high mean scores for

all areas – high scoring

population

• Your client’s score for Self-Insight was 77.

• This seems high in terms of a raw score

• The mean for the cohort is 75.29; SD 8.68 • (So the T score is going to be over 50, as the raw score is larger – but will

it be a high score or actually average?)

• Apply Z score formula: Z = (x – μ) Z = (77.50 – 75.29) = 2.21= 0.2546082*

σ 8.68 8.68

• Apply T score formula: T = (Z score x 10) + 50 so... (0.2546082 x 10) + 50

so... 2.546082 + 50 = 52.546082

FOR THE TABLE ROUND TO 2DP: Z = 0.25, T = 52.55 FOR THE CLIENT PROFILE ROUND TO 0DP: T = 53

• Your client’s standardised score is 53. Compared to the average British undergraduate, your client is actually in the average – they have the level of self- insight that would be expected for an educated person like a psychology undergraduate ... Which is good!

25 *IT IS IMPORTANT TO RETAIN A NEGATIVE Z SCORE IF YOU HAVE ONE AND RETAIN THE DECIMAL PLACES WHILST CALCULATING

AGAIN AN EXAMPLE – Make sure you work through this systematically

AGAIN AN EXAMPLE – Make sure you work through this systematically

• Your client’s score for Self-Insight was 77.

• This seems high in terms of a raw score

• The mean for the cohort is 75.29; SD 8.68 • (So the T score is going to be over 50, as the raw score is larger –

but will it be a high score or actually average?)

• Apply Z score formula:

Z = (x – μ) Z = (77 – 75.29) = 1.71= 0.1970*

σ 8.68 8.68

• Apply T score formula:

T = (Z score x 10) + 50 so... (0.1970 x 10) + 50

so... 1.970 + 50 = 51.97 = 51.97 (this is the figure for your table, 2DP)

• Your client’s standardised score is 52 (this is what you tell your client) • Compared to the average British undergraduate, your client is actually in the

average – they have the level of self-insight that would be expected for an educated person like a psychology undergraduate ... Which is good!

26 *IT IS IMPORTANT TO RETAIN A NEGATIVE Z SCORE IF YOU HAVE ONE!

**WECQ CALCULATIONS **

• Self Insight

• Mean: 75.29 SD: 8.68

• Communication

• Mean: 76.75 SD: 8.55

• Judgement

• Mean: 78.066 SD: 7.88

• Resilience

• Mean: 67.24 SD: 10.02

• Foundations

• Mean: 69.90 SD: 10.92

• TOTAL EIQ

• Mean: 73.44 SD: 9.21 27

Calculating your Z score

Z = (x – μ)

σ

Calculating your T score

T = (Z score x 10) + 50

Where x = score

U = mean

O = standard deviation (SD)

Scores are interpreted according to their range, not the numerical value

28

0 25 50 75 100

• Because psychology undergraduates are generally scoring quite high for

Emotional Intelligence, these principles will apply:

• A high score will be moderated because most undergraduates also score

highly

• A low score will become more extreme for the same reason 28

⚫ We are interested in what RANGE a score falls into in psychometrics

⚫ A score of 45-55 is within average range, 40-45 low average, 40-30 low,

below 30 very low; 55-60 is high average range, 60-65 high, 65-70 very high

and 70-80 excellent in comparison with your peers.

Mean score, Average range... 2 9

MEAN High Average

Low Average

Average Average!

All these scores

are within the

average range

WRITING UP WECQ RESULTS – SAME AS FOR PERSONALITY

• Introduce your table • The table should have a title which indicates that

is shows the descriptive statistics of your client’s scores.

• Give a brief comment on the content of the table. • Remember the standardised scores show your client’s scores as

they relate to the appropriate norms. • In this case, the norm for British undergraduates.

30

Because scores have been ‘normalised’ by the process of conversion to T scores, the mean score is now 50 = any score

around 50 means you are AVERAGE in that category = average for a British undergraduate, that is = what is expected for someone of

your educational level = GOOD EIQ!

WECQ AND 5 COMPETENCY SCORES

31

WECQ

DOMAINS

Raw

score

Sample Statistics

(n=561)

Z score T score

Mean SD

Self-Insight 75.29 8.68

Communication

Judgement

Resilience

Foundations

TOTAL Emotional

Competency

NEXT WEEK

• Week 18 – Discussing EQ and client

profiling – BRING YOUR NORMALISED

SCORES and SCORES FOR OTHER

ITEMS

32