6.1 Preparing your data for analysis

This can be a very time consuming experience and may also involve making decisions which will have an impact on the data analysis stage, such as whether you want to prepare and analyse all data or be selective. This will depend, in part on your research questions, but can also, necessarily, be determined by more pragmatic issues such as time and resources.

Preparing quantitative data (questionnaires)

Depending on how you choose to administer your questionnaires, you will have more or less preparation to do before you can start to analyse the data.

Paper questionnaires are quite rare now, but if, for whatever reason, you decide to use a paper questionnaire, you will first need to input the responses into a form which can then be analysed. Probably the most common way of doing this is to input the responses directly into a spreadsheet, such as Excel. This can then be used for simple analysis or the data can then be exported into software such as SPSS or SAS.

If you are using an online survey tool, such as SurveyMonkey, the the responses are directly exported into a suitable format, often an Excel cvs file. This can then be analysed in Excel, SPSS or SAS.

Preparing qualitative oral data

Analysing oral data, such as interview data, is nigh on impossible without first transcribing it into a written format.

However, transcription is a hugely lengthy process and while each interview may take only half an hour, a full transcrption could easily take two or three hours to produce. A lot of transcription is now done by professional transcribers. This takes much of the work out of the job but may still require careful checking. It is also very expensive and so may not be practical for practioner researchers.

The first decision you need to make, therefore, is whether you require a full or partial transcption. A full transcription allows to you analyse all the data collected and so may lead to some unexpected fidings. However, it is very time consuming. As such, some researchers prefer to partially transcribe interview data. However, this requires making some careful decisions about which bits of the interview to transcribe and which not. Much will depend on the research questions but you will still have to spend a substantial amount of time listening to the data to be in a position to make such a decsition.

The second decision is whether the transcription should include pauses and repeated words, etc. Such a high level of detail is absolutely necessary for conversational analysis but may not be required for research on student attitudes to ICT. Again, a lot depends on both the purpose of the research and the level of the analysis required.

After a transcription has been produced, the next decision is whether you wish to analyse it manually or to export it into a qualitative analysis tool, such as NVivo or AtlasTI. More information on these different analysis tools can be found in the next section.

Preparing qualitative written data

Written data can come from a number of different sources, such as fieldnotes, diaries, student materials and policy policy documents.

As with oral data, the first decision to make is whether you wish to analyse it manually or using qualitative analysis software such as NVivo and AtlasTI. This decision would depend on the amount of data to be analysed and the level of analysis required. A student diary or feedback on a student assignment may be easier to analyse manually because they either contain a relatively small amount of data or are too complex to export electronically. Larger amounts of wriiten data, such as policy documents, amy well lend them themselves more easily to analysis software.

Preparing multimodal data

Multimodal data is data which constitutes different forms of communication, such as words, sounds and visual images. Rather than seeing each mode of communication as isolated from each other, multimodality is interested in the interaction between these different modes of communication. An obvious example is a website, where the written text and visual images contribute to the overall content of the website, along with, in some cases, videos, podcasts or sound files.

As a result, traditional ways of preparing data, such as transcribing, may not be suitable as part of the rich picture of the data will be lost.

Last modified: Tuesday, 4 Mar 2014, 16:39