De Staic Research is happy to offer support at any stage of the research process, whether you are still trying to decide what the focus of the research is, are in the process of data collection or are analysing your research data – we can help, we can add value and we can save you valuable time and help you to avoid costly or embarrassing mistakes.
Clarifying your research questions
What exactly are you trying to find out? Many research projects fail to satisfy the purpose for which they were carried out because insufficient time and thought has been given to thinking through the core issues to be investigated and (in the case of surveys) this often leads to confused or unclear questionnaires which do not collect the appropriate data. With our vast experience in research we can help you quickly identify what the focus of your research is, and therefore to ensure that you make the best choices in terms of the wording of questionnaires and more generally, the choice of research design and methodology.
Designing your research project
Using an appropriate research design is essential to getting the answers you need.
Online surveys are easy to carry out and may often be a cheap and convenient method but because tools such as Surveymonkey or Qualtrics are so affordable and convenient this can mean that they are sometimes used in contexts which are not appropriate.
Many people often fail to think about sampling at all. Your approach to sampling will have a huge impact on your findings – often people use the people closest at hand: friends, family or immediate colleagues – but these people are unlikely to be typical of the group you are interested in and so the findings may not be representative. We can help you approach sampling in an intelligent way which will make sure that your research is valid and robust. Response rate is another important factor – we can advise how to manage this to avoid low response rates which again may threaten the validity of your research.
Surveys are great for producing quantitative summaries of the views of groups of people but care needs to be given to the format of questions (single choice rating scale, multiple choice, ranking, open ended) but they are not good for researching very complex issues, where qualitative or mixed methods research may be more appropriate.
Qualitative research is often the best choice when the issue you are investigating is complex and you wish to explore or understand it rather than simply measure how many people have certain views (as in a survey). Qualitative research most commonly involves methods such as focus groups, or semi structured interviews but can also include unstructured interviews or various types of observation. Qualitative methods can also include self completion methods such as participants filling in reflective diaries at regular intervals, regarding, for example. some aspect of a serve or product which is being evaluated. Sampling is less important in qualitative research than it is in quantitative research, since one is not trying to necessarily get a representative sample, but there are stills some important considerations in choice of sample such as maximum variation (getting as wide a range of participants as possible).
Mixed methods research – the best of both worlds
Using a combination of quantitative and qualitative methods can greatly increase the validity of a research project. Focus groups can be used to develop good survey questions or to further investigate patterns of response in a survey (for example a survey may have found that women were much less satisfied than men with their jobs, but it might not tell us why). Allowing respondents to talk in their own words about their attitudes to a product, service or issue can often be extremely revealing and will often give explanations and insights that a survey would not.
Developing data collection instruments
Developing the questions for a questionnaire needs careful consideration in terms of question wording, format and how they link together (the survey logic). Questions should be short, avoid ambiguous language, jargon or slang, should avoid leading or biasing the respondent towards any particular type of response and should only ask respondents things that they can meaningfully respond to in a questionnaire. For example, ”how are businesses in the UK changing”? is a very broad question which might require a long and complex answer (making it unsuitable for a questionnaire). Also, it is leading – it presumes that businesses in the UK are changing. It would be much more effective to consider what aspects of change one is really investigating and to develop specific questions on those aspects (these might include cost consciousnesses, approach to the environment, attitudes to outsourcing, use of online retail platforms, strategies for scaling) which might be turned into Likert scale items (with response options from strongly disagree to strongly agree).
Where possible. questionnaires should be piloted (trialled with a small sample of respondents to identify any problems or ineffectual questions) but if this not possible De Staic Research can intensively assess the questionnaire, using our professional experience and judgement to identify such issues.
One of the most common mistakes in research is to rush to data collection without having adequately developed the questionnaire. This can lead to disappointing or confusing results.
Data collection starts when you ‘go live’ with your project in terms of getting back your first completed questionnaire or finishing your first interview or focus group. (e.g. setting up online surveys or telephone interviews). De Staic Research can carry out a survey for you (online, telephone interviews, face to face) or conduct focus groups or interviews. With surveys, the important aspect of data collection ate delivering the survey to an appropriate sample and getting the highest possible response rate, through careful design of the questionnaire (making it as short as possible, with a clear and attractive design and most importantly having questions that make sense to the participants and which therefore engage them) and invitation emails (e.g. personalised email to named individuals rather than generic weblink) and through appropriate use of survey reminders (which can be set up to automatically go to all those who have not responded within a given time period). For some surveys, the use of incentives (such as amazon vouchers or entry into a prize draw) may be needed to ensure an adequate level of response. There is no magic number for survey responses; it depends on the type of analysis you need to do and the level of error that one is willing to accept (this is referred to as sample power). A good rule of thumb is to try to have a minimum of 100 survey responses, but remember that it really is not just about the sample size – how the sampled was chosen (the sampling strategy) is actually more important in determining the validity of your survey data. For example – if you want to predict how people will vote at the next election, would you get a more accurate result from a) a random sample of 100 people across the country or b) from 1,000 people that you stop in your local shopping centre? Almost certainly, the smaller random sample (a) would give a more accurate prediction, because it would not be biased by geographical location, and it would be more similar to the UK population on the basis of gender, age and ethnicity than a convenience sample in a shopping centre.
De Staic Research offers a full range of quantitative data analysis using SPSS. These include descriptive statistics (frequency tables, percentages, means, measures so central tendency and dispersion, factor analysis (extracting the key aspects of a number of variables and so reducing the number of variables), correlation, multiple linear regression (understanding which variables are the best predictors of an output or dependent variable). Many other techniques (e.g. cluster analysis) and tests (t-tests, chi square, Kruskal Wallis) are available depending on the nature of your data and the focus of your research.
Qualitative data analysis is usually carried out on transcriptions of focus groups or interviews, through it can also utilise audio, video, photography or documentary analysis. There are various approaches to qualitative data analysis (Thematic analysis, template analysis, content analysis, grounded theory) which partly depend on your data and partly on the focus of your research. De Staic Research uses NVIVO and ATLAS software to assist with qualitative data analysis.
Writing up and reporting
We are academics and former academics who have published in peer-reviewed journals and have authored a great number of research reports for business, public sector and third sector clients. We pride ourselves on writing which is of an academic standard but always tailored to the particular audience, to ensure accessibility. We also believe in imaginative and informative graphics and tables, but without gimmicks. We will tailor reporting to your needs and agree the reports specification with you. We can also advise on the most appropriate channels of dissemination for your research depending on which target audiences you may wish to influence.