Agency Report: Reasoning Doesn't Decline with Age
Anyone who's ever had a 'senior moment' shares the common knowledge that memory declines as we age. Many studies have confirmed this belief, and so far, TidePool users have seen the same result. The graph below shows the average scores for different age groups using the TidePool app.
It might be tempting to assume that all cognitive functions get worse with age, but that isn't the case. As shown here, reasoning (thinking logically and solving problems), which involves using memory, doesn't decline significantly.
Studies have also shown that consistent brain training can improve cognitive functions for older adults, including memory in daily activities, though results for reasoning are even stronger.
Of course, data like these aren't a prediction of the future. Any one person can be better or worse at memory or reasoning tasks, regardless of age. The graph above shows a cross-section of the TidePool population at one moment in time, and the app hasn't been around long enough to show results for individuals over time. The world is just learning what brain training can do!
Tweet us @TidePool_co to let us know if these results ring true for you.
Data was analyzed from TidePool games played between 8/10/14 and 1/21/15. Players were included in the analysis if they had 1) reported their age and 2) played at least 3 games in both the memory and reasoning categories. Memory games included Codebreaker, Lockpicker and Defuse while reasoning games included Getaway, DeadDrop, Escape and Snooper. Raw scores on these games were converted to scaled scores by subtracting the population mean and dividing by the population standard deviation. The lines plotted here represent linear trends (least squares) for age by median scaled score per player.
We verified that memory scores were indeed more affected by age than reasoning scores using a general linear model. The model included age, scaled median memory scores, scaled median reasoning scores and an interaction term of age by cognitive factor. Residuals from a Weighted Least Squares regression, weighting points by the inverse of the square of player age, were indistinguishable from Gaussian white noise. We found a significant effect (coefficient estimate= -0.014, log(p) < -10) of the interaction term (age by cognitive factor) indicating that memory scores decline more steeply than reasoning scores with increasing age.
