It is now widely recognised that subjects including sociology, politics, social policy and human geography in the UK have developed a 'generic deficit' in QM skills. One contributing factor is the UK's comparative failure to teach applied maths well at school level leaving most recruits to university social science with limited numeracy skills. However this is compounded by indifference to good QM teaching in many undergraduate curricula, which leads students to associate 'critique' with qualitative methodology, or non empirical theory, so that the weakness in QM capacity is reproduced.

Reversing this vicious circle requires the imaginative use of secondary data in teaching: focusing on 'real' evidence about relevant substantive issues to keep students engaged. This develops skills that are both more pedagogically useful and practically relevant than those concerned with primary data collection. However, the complexity of 'real' evidence makes most available secondary data barely suitable for teaching, where good student learning requires simple, clear examples.

Drawing on experience from the UK ESRC Quantitative Methods Initiative, I suggest some ways in which some of these obstacles can be overcome and the need for more virtual collaboration to achieve this.