I started this blog series with a discussion on why I think this survey will be of great benefit to social enterprises (part 1). I also mentioned my dilemma between generating a sufficiently large sample size and the need for targeting when identifying social enterprise for the survey (part 2). Now I want to talk about a very practical challenge: how to make sure that organizations actually take the time and fill in the questionnaire?
How have my fellow researchers outside of Turkey approached this challenge?
International experience shows that the response rates of most social enterprise surveys range between 10 and 20 per cent only. For example, the 2010-2012 Mercator Social Entrepreneurship research team in Germany sent out 1,710 questionnaires and had a response of 208 valid questionnaires (15 per cent). In 2009, the Finding Australia’s Social Enterprise Survey FASES () sent out 4,000 and collected only 365 responses that is a response rate of less than 10 per cent. Some researchers probably attracted more respondents with an offer to add them to an industry online directory (e.g. FASES or the on-going Social Enterprise Census in the US) or to send out the findings prior to official publication. FASES even offered gift vouchers to buy products and services from the other survey respondents (e.g. FASES). Other factors that according this experience, positively or negatively influence the response rates are:
- The length of the survey (15 to 30 minutes, 30-45 questions seems reasonable in most countries but may be too long in other countries where organizations have weaker capacities),
- The response format (multiple choice or text based),
- The amount and type questions on financial data and other data perceived to be sensitive (or which may simply not be available in the requested format), and
- The technical means of how the survey is delivered e.g. in many countries respondents prefer a personal conversation to completing an online survey (and vice versa).
So what do we do in Turkey?
Assuming a response rate of 15% and aiming to receive at least, say, 60 completed questionnaires we would need to reach out to at least 400 enterprises – quite a challenge in a place like Turkey with such a small sector. On the other hand the number of 60 respondents appears small in comparison with surveys carried out in other countries. However, it is a beginning and it would be sufficient to identify some patterns and draw initial conclusions.
With a view to increase my response rate as much as possible we want to offer respondents to take them through the questionnaire by phone instead of asking them to fill in the survey online only. However, phone interviews are only possible with organizations for which we have been able to gather contact details, or those who approached us and shared their contact details. We are also offering to invite participants to the pre-launch of the event, to send them the results prior to the official publication and, possibly, offering drawing of cash or shopping vouchers amongst respondents as additional incentives. We are also considering using the unique opportunity of this survey to set up a social enterprise online directory managed, for example, with www.sosyalgirisim.org or another suitable platform. Many social enterprises would see an advantage to be added to such a directory and thus increase their visibility. An online directory would also increase transparency of the sector and allow for more targeted intervention. This would be outside the scope of my project, however, so there is need for wider consultation.
Has this approach worked? Well, I don’t know yet. I will share my experience in one of the next blogs of this series in a few weeks once we have the first results.