Skills for Learning

Broadly speaking, 'statistics' refers to methods, tools and techniques used to collect, organise and interpret data. The goal of statistics is to gain understanding from data. Therefore, you need to know how to:

• Produce data – for example, by handing out a questionnaire or doing an experiment
• Organise, summarise, present and analyse data
• Draw valid conclusions from findings

There are a number of statistical methods you can use to analyse data. Choosing an appropriate statistical method should follow naturally, however, from your research design. Therefore, you should think about data analysis at the early stages of your study design. You may need to consult a statistician for help with this.

Tips for working with statistical data

• Plan so that the data you get has a good chance of successfully tackling the research problem. This will involve reading literature on your subject, as well as on what makes a good study.
• To reach useful conclusions, you need to reduce uncertainties or 'noise'. Thus, you will need a sufficiently large data sample. A large sample will improve precision. However, this must be balanced against the 'costs' (time and money) of collection.
• Consider the logistics. Will there be problems in obtaining sufficient high-quality data? Think about accuracy, trustworthiness and completeness.
• Statistics are based on random samples. Consider whether your sample will be suited to this sort of analysis. Might there be biases to think about?
• How will you deal with missing values (any data that is not recorded for some reason)? These can result from gaps in a record or whole records being missed out.
• When analysing data, start by looking at each variable separately. Conduct initial/exploratory data analysis using graphical displays. Do this before looking at variables in conjunction or anything more complicated. This process can help locate errors in the data and also gives you a 'feel' for the data.
• Look out for patterns of 'missingness'. They are likely to alert you if there’s a problem. If the 'missingness' is not random, then it will have an impact on the results.
• Be vigilant and think through what you are doing at all times. Think critically. Statistics are not just mathematical tricks that a computer sorts out. Rather, analysing statistical data is a process that the human mind must interpret!

Top tips! Try inventing or generating the sort of data you might get and see if you can analyse it. Make sure that your process works before gathering actual data. Think what the output of an analytic procedure will look like before doing it for real.

(Note: it is actually difficult to generate realistic data. There are fraud-detection methods in place to identify data that has been fabricated. So, remember to get rid of your practice data before analysing the real stuff!)

Software packages can be used to analyse and present data. The most widely used ones are SPSS and NVivo.

SPSS

SPSS is a statistical-analysis and data-management package for quantitative data analysis. Click on ‘How do I install SPSS?’ to learn how to download SPSS to your personal device. SPSS can perform a wide variety of statistical procedures. Some examples are:

• Data management (i.e. creating subsets of data or transforming data).
• Summarising, describing or presenting data (i.e. mean, median and frequency).
• Looking at the distribution of data (i.e. standard deviation).
• Comparing groups for significant differences using parametric (i.e. t-test) and non-parametric (i.e. Chi-square) tests.
• Identifying significant relationships between variables (i.e. correlation).

NVivo

NVivo can be used for qualitative data analysis. It is suitable for use with a wide range of methodologies. Click on ‘How do I access NVivo’ to learn how to download NVivo to your personal device. NVivo supports grounded theory, survey data, case studies, focus groups, phenomenology, field research and action research.

NVivo can:

• Process data such as interview transcripts, literature or media extracts, and historical documents.
• Code data on screen and explore all coding and documents interactively.
• Rearrange, restructure, extend and edit text, coding and coding relationships.
• Search imported text for words, phrases or patterns, and automatically code the results.

Miles and Huberman (1994) point out that there are diverse approaches to qualitative research and analysis. They suggest, however, that it is possible to identify 'a fairly classic set of analytic moves arranged in sequence'. This involves:

1. Affixing codes to a set of field notes drawn from observation or interviews.
2. Noting reflections or other remarks in the margins.
3. Sorting/sifting through these materials to identify: a) similar phrases, relationships between variables, patterns and themes and b) distinct differences between subgroups and common sequences.
4. Isolating these patterns/processes and commonalties/differences. Then, taking them out to the field in the next wave of data collection.
5. Highlighting generalisations and relating them to your original research themes.
6. Taking the generalisations and analysing them in relation to theoretical perspectives.

        (Miles and Huberman, 1994.)

Patterns and generalisations are usually arrived at through a process of analytic induction (see above points 5 and 6). Qualitative analysis rarely involves statistical analysis of relationships between variables. Qualitative analysis aims to gain in-depth understanding of concepts, opinions or experiences.

There are a number of different ways of presenting and communicating information. The particular format you use is dependent upon the type of data generated from the methods you have employed.

Here are some appropriate ways of presenting information for different types of data:

Bar charts: These may be useful for comparing relative sizes. However, they tend to use a large amount of ink to display a relatively small amount of information. Consider a simple line chart as an alternative.

Pie charts: These have the benefit of indicating that the data must add up to 100%. However, they make it difficult for viewers to distinguish relative sizes, especially if two slices have a difference of less than 10%.

Other examples of presenting data in graphical form include line charts and scatter plots.

Qualitative data is more likely to be presented in text form. For example, using quotations from interviews or field diaries.

Summary

1. Plan ahead, thinking carefully about how you will analyse and present your data.
2. Think through possible restrictions to resources you may encounter and plan accordingly.
3. Find out about the different IT packages available for analysing your data and select the most appropriate.
4. If necessary, allow time to attend an introductory course on a particular computer package. You can book SPSS and NVivo workshops via MyHub.
5. Code your data appropriately, assigning conceptual or numerical codes as suitable.
6. Organise your data so it can be analysed and presented easily.
7. Choose the most suitable way of presenting your information, according to the type of data collected. This will allow your information to be understood and interpreted better.

Information sources are sometimes categorised as primary, secondary or tertiary sources depending on whether or not they are ‘original’ materials or data. For some research projects, you may need to use primary sources as well as secondary or tertiary sources. However the distinction between primary and secondary sources is not always clear and depends on the context. For example, a newspaper article might usually be categorised as a secondary source. But it could also be regarded as a primary source if it were an article giving a first-hand account of a historical event written close to the time it occurred.

Primary sources

Primary sources enable you to get as close to the actual event or research as possible. Primary sources are useful for learning about perceptions of an event at the time it occurred, and for getting the most up-to-date information about a topic. Examples include, diary entries, census data, original accounts of events, journal articles with original reports of research, letters, email or other correspondence, original manuscripts and archives, interviews, research data and reports, statistics, autobiographies, exhibitions, films, and artists' writings.

Secondary sources

Secondary sources interpret, evaluate or analyse primary information resources. Secondary sources are useful for providing background information on a topic, or for looking back at an event from a current perspective. Examples include, journal articles which review or interpret original findings, popular magazine articles commenting on more serious research, textbooks, dictionaries and encyclopaedias, commentaries, and biographies.

Tertiary sources

This term is not used a great deal and there is overlap between what might be considered a secondary source and a tertiary source. One definition is that a tertiary resource brings together secondary sources. Examples include, almanacs, fact books, bibliographies, dictionaries and encyclopaedias, directories, indexes and abstracts.