Construct Validity - 16PF & MBTI

Once we have ascertained the reliability of a test, we must address its validity. It is important to know that the constructs we are measuring are valid, that it is indeed measuring the characteristic it purports to measure. The following data demonstrates that the dimensions of the 15FQ™ are consistent with similar measures of personality.

Relationship between 15FQ™ & 16PF

Correlations between 15FQ™ and 16PF-C, 16PF-A and 16PF-5 <> .32 <> .55 <> .53

	16PF Labels	16PF-A	16PF-C	16PF-5	15FQ™ Labels
A	Warm	.37	.54	.63	Outgoing
C	Calm-stable	.55	.59	.69	Calm-stable
E	Assertive	.53	.45	.57	Assertive
F	Happy-go-lucky	.69	.61	.64	Enthusiastic
G	Conscientious	.43	.67	.29	Conscientious
H	Venturesome	.76	.72	.75	Socially Bold
I	Tender-minded	.58	.42	.73	Intuitive
L	Sceptical	.46	.46	.59	Suspicious
M	Imaginative	.45	.31	.72	Conceptual
N	Calculating	.17	.48	.15	Restrained
O	Apprehensive	.58	.70	.82	Self-doubting
Q1	Experimenting	.28	.35	.59	Radical
Q2	Self-sufficient	.65	.68	.78	Self-sufficient
Q3	Controlled	.20	.25	.32	Disciplined
Q4	Tense	.56	.54	.25	Tense-driven
	FG/MD	Distortion

The table above displays the correlations between the 15FQ™ scales and their 16PF counterparts. The form A and form 5 data is based on a sample of 110 and 108 delegates on Psytech International training courses, while the form C data is based on a sample of 135 undergraduate volunteers. The correlation coefficients range from 0.25 to 0.72 for Form C, 0.17 to 0.76 for Form A and .15 to .82 for Form 5, with median values of 0.53 and 0.54 and 0.63 respectively. There is some indication in the above table of a tendency for the more reliable 16PF-A scales to correlate more highly with their 15FQ™ counterparts. In other words those 16PF scales which show the largest correlation with the 15FQ™ tend to be those scales which are the most reliable. This is not particularly surprising, one can not expect two scales to correlate highly if one of the scales is unreliable, and hence contains much error variance. This is further corroborated by the generally higher correlations obtained with 16PF-5. Given the higher levels of reliability for Form 5, it is generally easier to interpret the correlations between 15FQ™ and 16PF-5. The table above shows that there is a high level of congruence (r > 0.6) between eight 16PF-5 factors: A, C, F, H, I, M, O, and Q2 and their corresponding 15FQ™ dimensions. This should be evaluated in the light of the level of congruence between 16PF-5 and its predecessor 16PF-A, for which only five pairs of equivalently labelled factors: F, H, I, O, Q4 correlate higher than 0.6.

However, it is also evident from this table that certain scales appear to be measuring different aspects of personality. 16PF-5 factors G, N, Q3 and Q4 correlate less than 0.3 (ranging from 0.15-0.29) with their 15FQ™ counterparts. Given that the low correlations can not be explained by low reliability on the part of either test, this suggests that these measures are likely to be assessing different aspects of personality. Examination of the respective scales confirms this. Factors G, Q3 and N have been revised substantially in the new edition of the 16PF (Russell M & Karol D 1994). In each case, these three factors as measured by 16PF-5 have become more narrowly focused. Factor G more narrowly measures attention to rules; Factor Q3 concentrates on perfectionism and Factor N on privateness. Factor Q4 appears to be measuring aspects of frustration tolerance which is only one aspect of the 15FQ™ equivalent. Once again, it should be noted that the level of equivalence between 16PF-5 and the previous edition 16PF-A is lower with Factors L, M, N and Q1 correlating below 0.21 with the same label pair (range: 0.15-0.21).

Correlations between 16PF-A and 16PF-5 (both UK& US) 15FQ™ <> 0.49 <> 0.50

	16PF Labels	16PF5 (US)	16PF5 (UK)	15FQ™	15FQ™ Labels
A	Warm	0.52	0.59	0.47	Outgoing
C	Calm-stable	0.54	0.57	0.56	Calm-stable
E	Assertive	0.55	0.55	0.63	Assertive
F	Happy-go-lucky	0.71	0.80	0.69	Enthusiastic
G	Conscientious	0.41	0.46	0.63	Conscientious
H	Venturesome	0.80	0.85	0.70	Socially Bold
I	Tender-minded	0.55	0.71	0.71	Intuitive
L	Sceptical	0.38	0.15	0.43	Suspicious
M	Imaginative	0.17	0.21	0.34	Conceptual
N	Calculating	0.26	0.19	0.26	Restrained
O	Apprehensive	0.67	0.60	0.68	Self-doubting
Q1	Experimenting	0.31	0.15	0.55	Radical
Q2	Self-sufficient	0.66	0.51	0.53	Self-sufficient
Q3	Controlled	0.42	0.52	0.41	Disciplined
Q4	Tense	0.48	0.60	0.66	Tense-driven
	FG/MD	0.55	Distortion

The table above provides the correlation coefficients obtained between the 16PF-A and its equivalent scale in both the UK and US form of 16PF-5 and the 15FQ™. The anglicization of 16PF-5 is described in Smith (1994) . As can clearly be seen, many of the 15FQ™ dimensions provide larger correlations with 16PF-A than they do the 16PF-5 factors. The average correlation between 16PF-A and 15FQ™ is larger than that between 16PF-A and 16PF-5. Given that the 16PF fifth edition is ostensibly not a new test but merely a revised, improved, version of the 16PF this table provides powerful evidence in favour of the claim that the 15FQ™ is measuring highly similar personality traits to those of the 16PF-A.

The Relationship between the 15FQ™ & the MBTI

A sample of 221 volunteers completed both the 15FQ™ and the MBTI as part of a test validation exercise. Table 12 below displays the significant correlations between the various 15FQ™ dimensions and the MBTI.

The next table demonstrates some interesting, and meaningful, patterns of correlations between the 15FQ™ dimensions and the types identified by the MBTI. The four major extravert dimensions of the 15FQ™ - outgoing, enthusiastic, socially bold and self-sufficient - all show large correlations with the introvert/extravert MBTI types. The sensation/intuition types correlate fairly strongly with logically relevant 15FQ™ dimensions - conventional/radical, practical/conceptual. Thinking/feeling types correlate relatively well with 15FQ™ temperamental/stable, factual/aesthetically sensitive and confident/self-doubting. Finally, MBTI judging/perception types show good correlations with several relevant 15FQ™ dimensions - cautious/enthusiastic, flexible/Conscientious, direct/restrained and informal/disciplined.

Significant Correlations Between the 15FQ™ and the MBTI (N=221)

15FQ™ Scale	E	I	S	N	T	F	J	P
Outgoing	.75	-.71
Calm-stable	.31	-.45			.37	-.30
Assertive	.40	-.38	-.40	.29
Enthusiastic	.60	-.60	-.43	.37			-.44	.41
Conscientious							.39	-.40
Socially Bold
Intuitive					.30	.30
Suspicious
Conceptual			-.49	.53
Restrained
Self-doubting					-.34	.27
Radical			-.50	.52			-.43	.46
Self-sufficient	-.70	.67
Disciplined			.35	-.37			.50	-.48
Tense-driven
Distortion