ANALOGICA 50 Statistics Report Alex Tolio - 2024 © 1. Formal Statistics 2. Reliability Measures 3. Correlations with Professional Tests 4. Official Norms 5. Scatterplot 6. Preliminary Factor Analysis Μ ean raw 17.6 Standard Deviation 9.52 Mean IQ 140 Skewness 0.41494 Kurtosis -0.65352 SEM 1.45 Lowest Score 3 Highest Score 39 1. Formal Statistics Cronbach’s α 0.949 True Cronbach’s α 0.935 Mcdonald’s ω 0.954 Guttman’s Lambda 6 0.929 Kuder-Richardson 20 0.950 Pearson's ρ 0.9024 Spearman's ρ 0.8914 Unsolved Items 28, 44, 46, 52 Bad Items 7, 34, 41, 43, 47 Effective Ceiling 43/52 True Ceiling 52/52 Discrimination 110 - 175 (SD 15) Total Sample 46 2. Reliability Scales ANALOGICA 50 Tests Mean IQ Μ ean Raw ρ N Psychometric Batteries 142 17.6 0.902 17 3. Correlation with Psychometric Batteries Tests used: Old SAT (V), GRE (V), VAT-R, CAIT (VCI), WISC (VCI), WAIS-IV (FSIQ), JCWS, IAW, SBV (VFR), WISC (FSIQ), MAT, WISC (FRI), JCCES, SC-ULTRA, Old SAT (M), CAIT (PRI), UNKNOWN MENSA TEST, OLD SAT (M), C-09, SBV (FSIQ), MAB ll, IST, RAPM, FRT-A. ANALOGICA 50 Tests Mean IQ Μ ean Test IQ ρ N Psychometric Batteries 139.161 140.742 0.9027 31 ANALOGICA30 & ANALOGICA 50 Raw IQ Raw IQ Raw IQ 4 115 18 140 32 162 5 120 19 142 33 163 6 122 20 143 34 165 7 124 21 145 35 167 8 125 22 147 36 168 9 127 23 148 37 170 10 128 24 150 38 171 11 130 25 151 39 173 12 131 26 153 40 174 13 133 27 154 41 176 14 134 28 156 42 177 15 136 29 157 43 177+ 16 137 30 159 44 177+ 17 139 31 160 45+ 177+ 4. Official Norms 5. Scatterplot 6. Preliminary Factor Analysis EFA - 2 factor solution credit to: https://compositator.com/ g loading= 0.787 g loading(SLODR)= 0.893 After the first factor analysis and attempt to “find” the g loading, some crucial changes were made. First, the data was NOT properly prepared well enough, and thus, there was severe model “overfitting”, since data was used “as is” directly to the CA file. In return, this caused a “depression” of properties, such as omega η It is important that one aims for the best model fit possible, to ensure values obtained are as reliable as possible. However, a greater sample is needed for further validation, and the data yet remains preliminary, but great. What has changed due to data preparation and trimming: Items that were unsolved were entirely removed. Items with ONLY 1 to 2 solver(s) were entirely removed. The following items were removed for the above reasons: Q25, Q28, Q29, Q32, Q40, Q43, Q44, Q45, Q46, Q52. After such; more reliable values were obtained, to a severe improvement in model fit - in the cost of the removal of 10 items.