The influence of training on spatial ability

In a former 'post' (from August 31st) we've talked about the need for evidence about the influence of cognitive ability training on achievement in math, reading and writing.  Studies show that spatial ability predicts achievement in science, technology, engineering and mathematics.  Wai, Lubinski and Benbow (2009) found, for example, that the spatial skills of people with a Bachelor's (or higher) degree in engineering are 1.58 standard deviations higher than those of the general population.

Does spatial ability training improve achievement in science, technology, engineering and math?  This meta-analysis:

The malleability of spatial skills: a meta analysis of training studies.  David H. Uttal, Nathaniel G. Meadow,Elizabeth Tipton, Linda L. Hand, Alison R. Alden,  Christopher Warren and Nora S. Newcombe.  Psychological Bulletin 2013, Vol. 139, No. 2, 352–402

http://longevity3.stanford.edu/brainhealth/files/2013/04/uttal-et-al2013_malleability-of-spatial-skills-meta-analysis.pdf

went half way to answer this question.  The researchers have meticulously checked the literature, to try to answer the following questions:  Are spatial skills malleable and can training improve them?  If so – is the improvement durable? Does the improvement transfer to other spatial tasks?  Which is most effective training method? In light of research showing that men outperform women on spatial tasks – can training in those tasks close the gap between men and women?  Does training have a different impact in different age groups?

Why did I write that the meta-analysis went half way?  Because most studies included in this meta-analysis (202 out of 206 studies) examined the effect of training on other spatial skills and not on achievement in science, technology, engineering, math or other achievement domains (reading, writing, geography etc).  The studies were done with non-clinical population.

The researchers describe the meta- analysis process in a very interesting and a (mostly) clear way.

The results:

Are spatial skills malleable and can training improve them? Yes, by almost half a standard deviation compared to the  pre-training level.  What's more: participants who had low spatial ability on the pretest, improved more after training than participants who had high spatial ability on the pretest.

 If so – is the improvement durable? Yes.  There was no difference in the amount of improvement when it was tested right after training, less than a week after training, or less than a month after training.

Does the improvement transfer to other spatial tasks? Yes.  Training led to an improvement of about half a standard deviation in new spatial tasks, even those that required different skills or representations than those required in the training tasks.

Which is the most effective training method?  There was no difference in efficiency and amount of improvement between training through video games, through a course or training on a specific task with strategic instructions for that task.

 In light of research showing that men outperform women on spatial tasks – can training in those tasks close the gap between men and women? Men outperform women in spatial processing tasks with the exception of object location memory ("where have I left my keys?") in which women sometimes perform better.  Men outperformed women in   this meta-analysis, too – both in the pre and in the posttest, both in the treatment and in the control groups.  There was no difference in the amount of improvement between men and women.  Following training, both sexes improved equally well.  Training did not close the gap between men and women.

 Does training have a different impact in different age groups? In this meta-analysis, all age groups improved equally well following training.