A common activity in school mathematics is to derive a function from a table of values. There are two general strategies one might adopt to determine the function f. First, one might try to determine a connection between a particular n and f(n), and then evaluate whether it holds for other values of n. An alternative strategy would be to look for a relationship between values of f(n), f(n1), f(n2), and so on. The first of these strategies entails predominantly horizontal shifts of attention, whereas the second involves mostly vertical shifts of attention. In this brief article, the authors report on a study which sought to answer two questions. First, are there individual differences in strategy choices when tackling such generalisation problems? And second, if so, are these differences associated with a participants' mathematical background, or their success at tackling these problems?