SILC Showcase

Showcase June 2012: The Malleability of Spatial Skills: A Meta-analysis of Training Studies

Share this article using our url:

 The Malleability of Spatial Skills: A Meta-analysis of Training Studies

David H. Uttal1, Nathaniel G. Meadow1, Elizabeth Tipton1, Linda L. Hand1, Alison R. Alden1, Christopher Warren1, and Nora S. Newcombe2 (PI)

1Northwestern University, 2Temple University


For the archival version of this research, and the preferred citation, see:

  • Uttal, D. H., Meadow, N. G., Tipton, E., Hand, L. L., Alden, A., Warren, C., & Newcombe, N. S. (2013). The malleability of spatial skills: A meta-analysis of training studies. Psychological Bulletin, 139(2), 352-402. [DOI]

Are spatial skills malleable? If so, how much can we expect them to improve with training? And is that improvement durable and transferable?

 What are the implications of the malleability of spatial skills for STEM disciplines?

Having good spatial skills strongly predicts achievement and attainment in science, technology, engineering, and mathematics fields (e.g., Shea, Lubinski, & Benbow, 2001; Wai, Lubinski, & Benbow, 2009). Improving spatial skills is therefore of both theoretical and practical importance. To determine whether and to what extent training and experience can improve these skills, we conducted a meta-analysis of 217 research studies (206 after excluding the outliers) investigating the magnitude, durability, and generalizability of training on spatial skills.

Meta-analyses analyze and present results as effect sizes, which allow for an assessment of the effects of an intervention despite differences in how outcomes are measured. Effect sizes express differences in terms of Standard Deviations. Thus an effect size of 0.5 would indicate that training led to an improvement of one-half of a standard deviation in the measured outcome (Cohen, 1992). The meta-analysis showed that spatial skills are indeed quite malleable. Excluding outliers, the overall mean weighted effect size relative to available controls was 0.47 (SE = 0.04, m = 206, k = 1,038). Spatial training, on average, improved performance by almost one-half of a standard deviation.

Moreover, in at least some cases, the effect of training endured over times. We coded the time delay from the end of training until the posttest was administered for each study. Some researchers administered the posttest immediately, while others waited a few days, a week or even a month to administer the posttest. The magnitude of the training effect was not affected by the length of the delay. This result suggests that training can endure. Of course, those studies that included lengthy delays might differ from those that included only short delays. For example, the studies with lengthy delays may have used stronger or more effective forms of training. Nevertheless, the results do indicate that at least in some situations training can last long enough to be educationally meaningful.

Finally we also examined whether training transfers to other tasks. This issue is particularly important because if the effects of training are confined to performance on tasks directly involved in the training procedure, it is unlikely that training spatial skills will lead to generalized performance improvements in the STEM disciplines. Not all studies included measures of transfer, but those that did showed effects that were as strong as the original training effects (g = 0.48, SE = 0.04, m = 170, k = 764). Thus, there was no decline in the effect of training when moving from the tasks that were directly trained to those that were less like the training task. This result is also very important for education because it suggests that the effects of training would not be limited only to those tasks that were included in the training. Further research is needed to determine whether the training could affect STEM learning, but there is some evidence that it could (e.g. Mix & Cheng, in press, Sorby & Baartmans, 2000).

The findings that spatial skills are malleable, durable and transferable suggests implementation of formal programs targeting spatial skills could aid students’ spatial ability. Prior research gives us a way to estimate the consequences of administering spatial training on a large scale in terms of producing STEM outcomes. Wai, Lubinski, Benbow, and Steiger (2010) have established that STEM professionals often have superior spatial skills, even after holding constant correlated abilities such as math and verbal skills. Using a nationally representative sample, Wai et al. (2009) found that the spatial skills of individuals who obtained at least a bachelor’s degree in engineering were 1.58 standard deviations greater than the general population .The very high level of spatial skills that seems to be required for success in engineering (and other STEM fields) is one important factor that limits the number of Americans who are able to become engineers (Wai et al., 2009, 2010) and thus contributes to the severe shortage of STEM workers in the United States.

In this meta-analysis, we have demonstrated that spatial skills can be improved. To put this finding in context, we asked how much difference this improvement would make in the number of students whose spatial skills meet or exceed the average level of engineers’ spatial skills. We calculated the expected percentage of individuals who would have a Z score of 1.58 before and after training. As shown in Figure 1, increasing the population level of spatial skills by 0.40 standard deviation (the most conservative estimate offered by our findings) would approximately double the number of people who would have levels of spatial abilities equal to or greater than that of current engineers.

Figure 1

In summary, our meta-analysis shows that spatial skills can be improved with training, and the improvement is durable and can be transferred to distinct tasks. Spatial skills are not usually taught in school, but doing so could potentially make a big difference. It could, for example, potentially double the number of students with the level of spatial ability observed in current engineers.


  • ♦ Cohen, J. (1992). Statistical power analysis. Current Directions in Psychological Science, 1(3), 98-101.
  • ♦ Mix, K. S. & Cheng, Y. L. (in press). The relation between space and math: Developmental and educational implications. In J. Benson (Ed.), Advances in Child Development and Behavior (vol. 42). Elsevier.
  • ♦ Shea, D. L., Lubinski, D., & Benbow, C. P. (2001). Importance of assessing spatial ability in intellectually talented young adolescents: A 20-year longitudinal study. Journal of Educational Psychology, 93(3), 604–614. doi:10.1037/0022-0663.93.3.604
  • ♦ Sorby, S. A., & Baartmans, B. J. (2000). The development and assessment of a course for enhancing the 3-D spatial visualization skills of first-year engineering students. Journal of Engineering Education, 301-307.
  • ♦ Wai, J., Lubinski, D., & Benbow, C. P. (2009). Spatial ability for STEM domains: Aligning over 50 years of cumulative psychological knowledge solidifies its importance. Journal of Educational Psychology, 101(4), 817–835. doi:10.1037/a0016127
  • ♦ Wai, J., Lubinski, D., Benbow, C. P., & Steiger, J. H. (2010). Accomplishment in science, technology, engineering, and mathematics (STEM) and its relation to STEM educational dose: A 25-year longitudinal study. Journal of Educational Psychology, 102(4), 860–871. doi:10.1037/a0019454
You are here: SILC Home Page SILC Showcase Showcase June 2012: The Malleability of Spatial Skills: A Meta-analysis of Training Studies