There are going to be differing judgments offered on this one! The problem, of course, is that we can't (and shouldn't) set up experimental tests of this question, so the evidence will all be correlational, and the door will remain wide open to differing interpretations.

Digit Span scores correlate positively with education, though the magnitude of the correlation is not, I think, very large - nothing like what one sees with, say, Vocabulary measures.

But does education increase working memory capacity? If so, is this a global boost, or limited to stimuli that resemble things we study in school (such as numbers)? Or is it simply that brighter people (with better working memories) tend to go further in school?

If you are concerned about the validity of Digit Span in your less-educated samples, you might consider using a different task. A grocery list or a list of errands might work. Of course, it may be less elegant and symmetrical than Digits Forward vs. Digits Backward.

Thank you for your answer. I understand the concern! I was mainly refering to uneducated or less educated immigrants wanting to learn a second language, that didin't have the choice whether to go to school or not. Especially women, sadly.

However, if I were to measure working memory capacity with the Digit Span Test on educated samples, and another (easier test) on less educated samples, that would already imply a difference in working memory capacity?

You are correct - it would not be appropriate to use separate working memory tests if you want to compare groups. And I share your concern about using Digit Span with people who have never been to school, especially if they have also been kept out of the marketplace. An uneducated street vendor or homemaker will often be very good with numbers (they need to keep track of money!), but someone who has been sheltered will have had less chance to develop the skills. For although we speak of working memory as a broad construct, performance on specific working memory tasks is highly susceptible to practice. (There was a famous case study in which a young man was induced to practice Digit Span tasks every day for months - he achieved a remarkable level of proficiency - but this did not generalize to other working memory tasks.)

I guess your options are (1) to find an alternative working memory task for all of your participants, which could be used in addition to or instead of Digit Span, or (2) to try for statistical control of the possible effect of education, as in partial correlations or analysis of covariance.

Thank you both for helpful answers! I will keep them in mind during my research

Hi Catrine. I am working on the same kind of question : to what extent literacy and schooling influence the development of working memory.

Studies on illiterate and unschooled people (for socio-economic reasons) have showed that poor verbal short-term memory is usually observed in illiterate adults, who display poor word span and digit span (e.g., Kosmidis, Zafiri, & Politimou, 2011; Morais et al, 1986; see Ardila et al., 2010, for a review). The origin of these effects may be partly attributed to lack of formal education or to illiteracy. The question is still open.

Also, beside all the cognitive consequences of attending school, schooling certainly develops a predisposition toward achievement in testing, as noted by Ardila et al. (2010). Some authors, like Kosmidis or R. Kolinsky and J. Morais, have sometimes used some tests more ecological or at least, more adapted for less educated people or illiterate adults (from Brazil, Portugal or Greece ). Their papers can be very interesting for you

I just started working on a similar question for my PhD. It is concerned with correlation between working memory and school achievement in illiterate adults who joined basic elementary education (specially designed for adults). I am using several WM tasks and Raven advanced matrixes as a test of intelligence. On the basis of previous discussion I guess I should test WM only in individuals just started the program, as individuals who are second and third year of the program already possess education that could confound results. I am also thinking to include another group (University students) to investigate their WM correlation to academic success, and further compare these two correlations (illiterate and university students of the same age range 19-24). I still have doubts whether this is a question ambitious enough for the thesis.

I am using following tasks:

1. Digit span (forward and backward) from Wechsler intelligence subtest numbers

2. Faces from Wechsler WMS III (recognition, short-term memory)

3. 2 – back task (with 4 x 4 matrixes, where one should remember last two positions of colored cell in a matrix)

4. I designed a complex task with animals where one should remember animals seen and simultaneously is inhibited with performing additional task.

I also wonder whether to include some more verbal WM task.

We no doubt see correlations between education and working memory. I work in genetics and had the opportunity to be a part of a working memory study with ~2300 participants. After looking at education and in thinking about genetic predisposition I have to wonder if people who are born with genes that help working memory function more efficiently have an easier time in school from the beginning. This could increase the chances that they are going to choose continuing education after high school. Although education might help to strengthen working memory in ways, the ability for strong working memory is already present.