INTRODUCTIONThe most fundamental question you need to ask yourself before thinking about, planning, or embarking on undertaking or contracting research is – what do you want to know and why?
This sounds a dumb question – but – a lot of folks don’t ask this. They just collect information for the sake of it or do what they have always done, analyse the same statistics in the same old way. Or perhaps people try and collect every available fact or capture all available data no matter how relevant. Statistics and research can tell you a great deal about your locality or region. There is a lot of information available these days, especially with the increase in availability of online data sources and research. However, one of the biggest mistakes to make is to rush out and gather as much information as possible, as you will soon be overwhelmed. Its not good to be ‘data driven’ in this way. It is much better to be needs, or demand driven. Gathering and interpreting data and information can take a lot of time and resources. Quite often, research or strategy department budgets are small. Therefore you need to ask some basic questions at the beginning.
RESEARCH MUST BE DEMAND- OR NEEDS- DRIVENTake a tip from a professional, experienced researcher and analyst – take some time to build a sound reason and rationale for a piece of work. Talk to people – especially the ones who are looking for intelligence or analysis to help them make a decision. Try and understand the context and needs from other people’s points of view. Remember – if you don’t supply them with intelligence or analysis that they can use, your work (and you) will not be valued.
Your needs: be clear about
– The research question – what do you want to find out?
– What’s it for? What use will be made of the answer?
– Who is it for? Is it for a specific group of people?
The ultimate value of research is to help people make better-informed decisions. Sometimes its hard to determine exactly what those needs are if other people are involved. Often they don’t know about research in a particular field to be able to articulate exactly what they want.
For example, a colleague might ask you to implement a ‘survey on business survival rates’. However, what they really want is to learn more about what they read in the paper this morning about their local area having much higher rates of business closure than the national average. They don’t know much about research or how its done but they assume most of it is concerned with performing surveys. So this leads them to ask you for one. Rather than doing exactly what you are asked or told to do its worth going back to this colleague and asking exactly what they want to know and why they want to know it, and also what kind of decision or actions will rest on the outcome of this research. If there is no decision or money resting on the research, then you have to ask yourself if it is a priority. Perhaps it could be a theme of your next annual report on your local economy, or perhaps you could dig around existing data and studies to get the answers rather than commission an expensive survey.
DATA IS RAW, INFORMATION IS COOKED, INTELLIGENCE IS FOUR COURSE MEALThe analagy is crude, but its true that data is raw information, and to be more meaningful it needs to be prepared and processed and ‘cooked’. However, you need to go one step further – to get intelligence out of the information – you need to analyse it. If you have cooked ingredients, you need to assemble and combine them into somehing palatable and digestable. It’s the same with information – you need to combine and analyse it so it tells you something that is digestible and meaningful. The food analogy is useful though – for example, you sometimes don’t need to cook a four course meal when a snack will do! Or maybe you have been forced on a diet because of government budgetary cuts. I digress.
Back to reality now – I tend to think of the following steps as cooked, raw, cooked and a four-course meal:
• Raw: the basic data that is generated by surveys of companies, individuals or is collected by government agencies and the like. At this stage, information is collected systematically (I hope – more about that in Chapter X) and quality checked/controlled (bad data or mistakes are rooted out).
• Cooked: the basic data has been quality checked and approved and collated systematically – now information is generated in the form of tables and charts. Things like crosstabulation and statistical tests or calculations can be performed. Quite often we use standard definitions for measuring phenomenon such as ‘unemployment’, ‘employment’, or ‘qualification levels’. The data is compiled according to these standard definitions. There may or may not be some descriptive information about the data.
• A four course meal: the tables are analysed to find out what they mean! This is a bit more complex than it sounds, but if we were looking at tables and charts of information concerning our local economy we might be looking at issues such as – how different is our economic structure in terms of industries to the UK average structure? or how high is unemployment here compared to the national average? We might find that high unemployment exists alongside high levels of job vacancies – we might ask what that means, and look at other data tables (such as migration statistics) to find out. Much analysis is about ‘triangulation’ – which is basically the cross-examination of different sets of information to ascertain a confident picture of what is occurring in the economy. Much of analysis is about gathering a set of incomplete pieces of information about an economy and seeing if any of these pieces fit together to give a picture of what is occurring. Or its like uncovering various clues about a crime. Some clues are circumstantial, some are red herrings, but some might fit together quite logically.
Of course there are some consultancies out there who could offer you a bit of fast food! But beware, too much fast food doesn’t do you good in the long run!
I have seen far too many descriptive reports incorrectly described as an ‘analysis’. For me, an analysis asks what the data means and describes the strength or association of causal factors. Merely saying that X is higher than Y doesn’t cut it as analysis for me.
