Statistics of Analogies subtest of Long Test For Genius

Norms (G = # right):

G Protonorm
0 187 or lower
1 207
2 227
3 253
4 273
5 280
6 320
7 347
8 360
9 367
10 380
11 386
12 388
13 400
14 420
15 433
16 440
17 447
18 460
19 500
20 567
21 573
22 607
23 630
24 647
25 667
26 687
27 707
28 727
29 747
30 767
31 787
32 807
33 827
34 847
35 867
36 887
37 907
38 927
39 947

• Quality of norms: 1.00

Scores so far

Analogies, © 1995, subtest of long Test For Genius (English version), has drawn these scores out of 39 (65 scores):

0*
4*
5*
6***** *
7**
8****
9***** *
10*****
11***
12***** *
13***
14***** *
15***
16****
17****
18***
19***
21*
22**
24*

Remarkably (for high-range tests), the top scorer is a female.

Correlation with other tests:

Mean correlation: 0.4

Used for norming are tests correlating .77 or higher. Prior scores are converted to the same scale (SD=15) if needed [and in March 2008 converted to protonorms using the initial protonorm formula]. Raw and prior scores are put in columns of numerical order, the highest facing the highest etc. Each raw score is normed at the median of the IQ's facing it. Scores above 24 are extrapolated.

Addition March 2008

Remark: This was a subtest of the no longer used Long Test For Genius, and most of its problems now occur in the Test For Genius - Revision 2004.

Scores

• # candidates: 98
• Median: 12.0
• Quartile deviation: 3.5
• Range: 24
• Maximum possible: 39
• Mean: 12.4
• Standard deviation: 4.66
• Male mean: 12.6
• Female mean: 10.6 (# females: 8)
• Hardness: 0.68
• Resolution: 0.19

Scores by males

n = 89

5**
6********
7**
8******
9*******
10********
11******
12*******
13*******
14*******
15***
16******
17******
18****
19******
21*
22**
23*

Scores by females

n = 8

0*
4*
10**
12**
13*
24*

Correlation with other tests

Weighted average of correlations: 0.545

Conservatively estimated minimum g loading: 0.74

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 to obtain the true, honest statistics; correlations are not artificially inflated by leaving out ceiling scores, outliers or other anomalies.

National averages

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

Correlation with personal details

Correlation with personal details - within females

Correlation with national IQs

Correlation of this test with national average IQs published by Lynn and Vanhanen:

• r = 0.07
• n = 79

Reliability and standard error

Using the "split-half" method:

• Rel = 0.79
• SE = 2.1 raw score points

Scores by age

Scores by year taken

Correlation year taken vs average score = 0.62
n = 96

Robustness and overall test quality

• Rob = 0.38
• Quality = 0.68

Item analysis

Item statistics are not published as that would help future candidates. To detect bad items, answers and comments from candidates are studied, as well as, for each problem, the correlation with total score (validity of the item in this test) and the proportion of candidates getting it wrong (hardness of the item). Possible bad items are removed or revised, resulting in a revised version of the test.