Statistics of behaviour scores

Introduction

Over the past twenty years I have been collecting various kinds of data for thousands of test candidates. This report deals with the predictive validity of a number of variables with regard to certain types of objectively observable misbehaviour that, sadly, occur in a small fraction of the candidates.

Insults, threats, and (re)tests under false names are common responses of these people when they are disappointed about their scores or otherwise dismayed. Notice that the false names are an implicit insult, as they result from their thinking "if he sees my real name he will deliberately give me a too low score to make it match my earlier score(s) on his tests or because he does not like me".

The fact they use this reasoning reveals their own dishonesty; this thought can only occur to them because it reflects what they would do in my place, or, as one says, "Ill doers are ill deemers".

To correlate misbehaviour to other statistics I added a field that contains 0 in the case of misbehavers and 1 in the case of people with objectively observable markedly positive, constructive behaviour. I left it blank in all other cases. Leaving these "neutral" cases blank rather than giving them, for instance, .5, makes this study more sensitive. The following statistics result:

Scores on Observed behaviour

Contents type: Assessment.   

• n: 133
• Mean: 0.50
• σ: 0.50
• Range: 1.0
• Maximum possible: 1
• Male mean: 0.43 (n = 113)
• Female mean: 0.90 (n = 20)
• Hardness: 0.50
• Resolution: 0.12

Scores by males

n = 113

Scores by females

n = 20

Correlation of Observed behaviour with other tests by Paul Cooijmans

Weighted average of correlations: 0.301

Conservatively estimated minimum g loading: 0.55

Ranking in above table is based on the unrounded correlations. All available data is present in this table, no tests are left out except for those with less than 5 score pairs. All known pairs are used, including possible floor/ceiling scores or outliers.

Note: One should realize this is a point-biserial correlation (where one of the variables is a dichotomy, which through its limited variance suppresses the size of the correlation) so the actual g loading of positive, constructive behaviour combined with the reverse of insulting behaviour may be higher than shown above.

Correlation of Observed behaviour with tests by others

Weighted average of correlations: -0.095

Ranking in above table is based on the unrounded correlations. All available data is present in this table, no tests are left out except for those with less than 3 score pairs. All known pairs are used, including possible floor/ceiling scores or outliers.

Please be aware that correlations with these external tests are in most cases affected (depressed, typically) by one or more of the following: (1) Little overlap with the object test because of the much lower ceilings and inherent ceiling effects of the tests used in regular psychology; (2) Candidates reporting scores selectively, for instance only the higher ones while withholding lower ones; (3) Candidates reporting, or having been reported by psychometricians, incorrect scores.

Remark: The non-positive correlation with tests by others may indicate that insulting, misbehaving persons are (even) more dishonest in reporting their other scores than are candidates in general; for instance, more likely to report only their highest scores, or to report incorrect high scores.

Correlation of Observed behaviour with other tests by Paul Cooijmans - for females

Weighted average of correlations: 0.303

Conservatively estimated minimum g loading among females: 0.55

Ranking in above table is based on the unrounded correlations. All available data is present in this table, no tests are left out except for those with less than 3 score pairs. All known pairs are used, including possible floor/ceiling scores or outliers.

Estimated loadings of Observed behaviour on particular item types

These are estimated g factor loadings, but against homogeneous tests (containing only particular item types) as opposed to non-compound heterogeneous tests. Although tending to surprise the lay person, it is not uncommon for tests to have high loadings on item types they do not actually contain themselves. Such loadings reflect the empirical fact that most tests for mental abilities measure primarily g, regardless of their contents; that the major part of test score variance is caused by g, and only a minor part by factors germane to particular item types. It is of key importance to understand that this is a fact of nature, a natural phenomenon, and not something that was built into the tests by the test constructors.

Compound tests have been left out of this table to avoid overlap.

Balanced g loading = 0.47

National means for Observed behaviour

For reasons of privacy, only countries with 2 or more candidates are included in this table. Ranking is based on the unrounded means, and then alphabetic.

Correlation with national I.Q.'s of Observed behaviour

Correlation of this test with national average I.Q.'s published by Lynn and Vanhanen:

• r = 0.11 (n = 132)

Correlation of Observed behaviour with personal details

P.S.I.A. True and Ethics (the latter combines True with the reverse of Cruel) are good at identifying misbehavers, while Extreme, Cruel, and Antisocial also help. But the question is if the P.S.I.A. would still work this well if explicitly used for selection; candidates might try to manipulate the outcome by answering dishonestly.

Interesting newcomers are Cooijmans Inventory of Neo-Marxist Attitudes and Candidate's self-estimated I.Q., but their n is too low to make these promising correlations with observed behaviour significant. Observed associative horizon is an obvious good correlate of behaviour, but as it concerns an assessment it is of less practical value than an actual personality test score.

Correlation with personal details of Observed behaviour - within females