Statistics of NIT-Form 1 and its subtests

NIT-Form 1, developed around 2012, was one of the predecessors of NAIT and remained in use for approximately 3 years.

It was an untimed, 115-item heterogeneous test that included numerical, verbal, logical, abstract, and spatial questions.

During that time, it was accepted as an admission tool by several high-IQ societies, including the SPIQR Society, the Real IQ Society, the Global Genius Generation Group, the High Intellect Society, and WIQF.

Below you can find the final statistics from the test before it was officially discontinued.


Statistical Summary
Test or subtest N Lowest raw score Highest raw score Mean Median Raw score standard deviation Quartile deviation Skewness Excess Kurtosis Cronbach's alpha Split-half reliability index Spearman–Brown corrected
NIT-Form 14445104.576.9776.513.678.81-0.28-0.080.920.870.93
Abstract1013.521.515.9516.53.32-0.781.040.760.640.78
Numerical921.52317.7418.53.661.75-1.894.690.840.740.85
Verbal6462417.8917.53.812.63-0.640.210.790.670.81
Logical6532214.37153.842.5-0.610.260.790.670.80
Spatial840.52010.05104.1230.11-0.370.80.680.81
Ivec's Quality
Test or subtest Number of items Good items Average Items Suspicious items Bad items Unsolved items Overall quality (%)
NIT-Form 1115842344078.1
Abstract2219210083.2
Numerical2320201086.0
Verbal2620411079.2
Logical2215610078.3
Spatial2210912065.3
More Statistics
Test or subtest Range Resolution Sample-dependent hardness Sample-independent hardness Standard error of measurement
NIT-Form 160.518.880.330.363.78
Abstract194.250.250.271.63
Numerical22.53.50.20.141.45
Verbal192.880.330.311.74
Logical2030.320.401.75
Spatial20.560.550.651.84
Correlations matrix between subtests
Abstract Numerical Verbal Spatial Logical
Abstract10.6660.6700.5270.474
Numerical0.66610.6850.7850.755
Verbal0.6700.68510.7320.779
Spatial0.5270.7850.73210.380
Logical0.4740.7550.7790.3801
g-loadings
Numerical subtest0.920
Verbal subtest0.914
Spatial subtest0.812
Logical subtest0.803
Abstract subtest0.783

Variance explained by g (first principal component): 0.720, meaning 72% of the total variance in subtests is accounted for by the general factor.

Method used for computing g-loadings: Principal Component Analysis (PCA) applied to the subtest correlation matrix; the first unrotated principal component was interpreted as g.

Back to main page