Validity of estimating IQ using chimp test and chess ELO

[drumbo]

Working memory has been found to be strongly predictive of intelligence. From Schneider and Niklas "Intelligence and Verbal Short-Term Memory/Working Memory: Their Interrelationships from Childhood to Young Adulthood and Their Impact on Academic Achievement" (2017):

Quote

Although recent developmental studies exploring the predictive power of intelligence and working memory (WM) for educational achievement in children have provided evidence for the importance of both variables, findings concerning the relative impact of IQ and WM on achievement have been inconsistent. Whereas IQ has been identified as the major predictor variable in a few studies, results from several other developmental investigations suggest thatWMmay be the stronger predictor of academic achievement. In the present study, data from the Munich Longitudinal Study on the Genesis of Individual Competencies (LOGIC) were used to explore this issue further. The secondary data analysis included data from about 200 participants whose IQ and WM was first assessed at the age of six and repeatedly measured until the ages of 18 and 23. Measures of reading, spelling, and math were also repeatedly assessed for this age range. Both regression analyses based on observed variables and latent variable structural equation modeling (SEM) were carried out to explore whether the predictive power of IQ and WM would differ as a function of time point of measurement (i.e., early vs. late assessment). As a main result of various regression analyses, IQ and WM turned out to be reliable predictors of academic achievement, both in early and later developmental stages when previous domain knowledge was not included as additional predictor. The latter variable accounted for most of the variance in more comprehensive regression models, reducing the impact of both IQ and WM considerably. Findings from SEM analyses basically confirmed this outcome, indicating IQ impacts on educational achievement in the early phase, and illustrating the strong additional impact of previous domain knowledge on achievement at later stages of development.

The chimp test is known to measure working memory. I tried the chimp test on humanbenchmark and got a score of 13 which corresponded to the 88.7th percentile:

I figure people who have played chess for a long time can also use their ELO scores as a measurement of how smart they are. I have been playing chess for 13 years and my ELO on LiChess corresponds to the 77th percentile:

Averaging these two percentiles gives 82.85%, which corresponds to an IQ of 114-115 assuming a normal distribution with a mean of 100 and a standard deviation of 15. This seems quite accurate to me based on my personal experiences. What do you think about the validity of measuring your IQ in this way?

Edit: A bit more information to help diagnose the cause of my poor visual memory; I tried the full scale IQ test on the openpsychometrics website and did well on the spatial portion:

How can we explain the inconsistency between my poor performance on the visual memory test and my strong performance on the spatial test, wouldn't those be related?

Edited September 12, 2020 by drumbo
Additional data

[CharonY]

1 hour ago, drumbo said:

How can we explain the inconsistency between my poor performance on the visual memory test and my strong performance on the spatial test, wouldn't those be related?

Without reading the rest, no generally not. There is a large body of research on it, which is outside me realm of expertise. However, at least since the 2000s folks have fond that spatial and visual memory can be disrupted by differently using different cues. These suggest that both forms of memory are processed separately. I am not sure whether since then folks have identified the neurological foundation for this separation.