To stimulate social science research in academia, the Higher Education Commission (HEC) has announced establishing Social Sciences Research Council later this year. Moreover, Rs90 million have been pumped into the current Social Sciences Thematic Research Program for the fiscal year 2015-16.

Establishing an independent council for quality social sciences research sounds like a good idea. But will the new council be able to pull off what the current Social Sciences and Humanities Committee has been unable to accomplish for more than a decade? Unless the HEC devises a comprehensive research strategy, throwing money at the problem will only result in more of the same old. To this end, any strategy will have to address both policy and university level issues.

At the policy level, the council will have to develop and refresh a research agenda keeping in view the kinds of problems that need addressing. To ensure relevance, research agenda setting committee should engage academics, members of the policy community, among others. The current mechanisms for reviewing research proposals have to be improved as well. Reviews should have a quick turnaround so that research can be initiated in a timely fashion. Faculty researchers often lament that grant application process is unreasonably long and tedious. Identification of qualified reviewers is another issue that needs tackling to ensure quality.

At the university level, the HEC will have to solicit university’s commitment to becoming centres for research excellence. This entails, among other things, re-evaluating research methodology courses with support from the HEC. In order to prepare social scientists equipped with skills to do meaningful research, universities should ensure that their graduate students learn to identify problems worth researching. They should also teach at least one quantitative analytical tool such as the Statistical Analysis System (SAS) and/or R -- a statistical software. The skill to translate research in multiple formats for diverse stakeholders is another imperative skill.

Ideally, graduate courses should teach methods using publicly available large datasets. University researchers in social sciences are often under the impression that good research implies collecting one’s own data. Ignoring secondary analysis of ‘big data’ is a huge missed opportunity for conducting good quality research. The word ‘big’ in big data refers to the "large" size of the data, which poses challenges to its analysis. Research based organisations outside academia have already set a precedent of producing high quality research using big data.

The Institute for Social and Policy Sciences (ISAPS) conducts policy relevant research in several areas of social sciences. In addition to conducting empirical research the organisation uses large datasets such as the NEMIS -- data pertaining to all educational institutions in both public and private sectors from pre-primary to tertiary levels; to produce relevant research for policy makers.

Other notable examples include Alif Ailaan’s report "Not Free At All" to profile costs borne by parents to send their children to both public and private schools. Using recent Pakistan Social and Living Standards Measurement (PSLM) data as well as data from another survey, they made recommendations that have direct relevance for policies around free education for all.

Similarly, Alif Ailaan, in partnership with the Society for Community Strengthening and Promotion of Education, Balochistan (SCSPEB), analysed matriculation examination results of 37,000 students in Balochistan to draw policy recommendations for quality education in the province.

Using large datasets, these organisations are utilising expensively collected data which otherwise goes under-utilised, making valid recommendations based on large samples, and generating research within a reasonable timeframe to address pressing issues.

Secondary data analysis has its limitations. Not having been designed keeping the needs of potential researchers in mind, one has to be very careful about identifying the types of questions that can be answered using secondary data set. Notwithstanding its limitations, analysis of secondary data is a powerful means of conducting relevant and rigorous policy research.

Realising the potential of big data to facilitate research, governments around the world now have websites dedicated to hosting data sets from various government departments, chief among them the US (http://www.data.gov/), the UK (https://data.gov.uk/), Canada (http://open.canada.ca/en) among others.

Large-scale data sets are increasingly becoming available in Pakistan but are not easily accessible in usable formats. Alif Ailaan and the Sustainable Development Policy Institute (SDPI) have taken the lead in Pakistan to make public data on education and nutrition available through the Pakistan Data Portal (www.data.org), at present hosting 134 datasets.

The HEC should create a data portal to share all publicly available datasets in usable format to facilitate data enthusiasts who will find innovative ways to apply available data.

Besides making resources available, the HEC should invest in outreach and skill development so that researchers know about and use these resources. Several such training models exist that can be replicated. For instance, the US National Center for Education Statistics (NCES) provides funding opportunities to researchers and students, graduate dissertation grants, as well as hands-on training to encourage use of large-scale datasets from federal agencies.

Another missed opportunity for conducting social science research is ignoring electronically available big data. This data can greatly influence the way social sciences research is approached.

Some early examples of uses of electronically available big data in social science included sentiment detection of social media posts made on Twitter and Facebook. There are now dozens of free and paid websites that perform sentiment detection on live streams of tweets matched against search words. A more involved example is the analysis of Twitter traffic to observe the effects of social unrest (post-Arab Spring, 2010) on the properties of information flows in social media.

Locally, researchers from the National University of Science and Technology, performed sentiment analysis of Pakistani elections 2013 using tweets and proved that the analysis matched, considerably, actual results published by Election Commission of Pakistan.

Though exciting things are happening in social science research, they are just not happening in academia despite large investments. The HEC will have to put in more thought into its programmes to be able to get value for money spent. At the moment there is little indication of that happening.