The data analysis process involves **three steps**: **(STEP ONE)** select the correct statistical tests to run on your data; **(STEP TWO)** prepare and analyse the data you have collected using a relevant statistics package; and **(STEP THREE)** interpret the findings properly so that you can write up your results (i.e., usually in Chapter Four: Results). The basic idea behind each of these steps is relatively straightforward, but the act of analysing your data (i.e., by selecting statistical tests, preparing your data and analysing it, and interpreting the findings from these tests) can be time consuming and challenging. We have tried to make this process as easy as possible by providing comprehensive, step-by-step guides in the Data Analysis part of Lærd Dissertation, but you should leave time at least one week to analyse your data.

Select the correct statistical tests to run on your data

It is common that dissertation students collect good data, but then report the wrong findings because of selecting the incorrect statistical tests to run in the first place. Selecting the correct statistical tests to perform on the data that you have collected will depend on **(a)** the research questions/hypotheses you have set, together with the research design you have adopted, and **(b)** the type and nature of your data:

The research questions/hypotheses you have set, together with the research design you have adopted

Your research questions/hypotheses and research design explain what variables you are measuring and how you plan to measure these variables. These highlight whether you want to

**(a)**predict a score or a membership of a group,**(b)**find out differences between groups or treatments, or**(c)**explore associations/relationships between variables. These different aims determine the statistical tests that**may**be appropriate to run on your data. We highlight the word**may**because the most appropriate test that is identified based on your research questions/hypotheses and research design can change depending on the type and nature of the data you collect; something we discuss next.The type and nature of the data you collected

Data is not all the same. As you will have identified by now, not all variables are measured in the same way; variables can be dichotomous, ordinal, or continuous. In addition, not all data is

**normal**, as term we explain the Data Analysis section, nor is the data you have collected when comparing groups necessarily**equal**for each group. As a result, you might think that running a particular statistical test is correct (e.g., a dependent t-test), based on the research questions/hypotheses you have set, but the data you have collected fails certain assumptions that are important to this statistical test (i.e.,**normality**and**homogeneity of variance**). As a result, you have to run another statistical test (e.g., a Mann-Whitney U instead of a dependent t-test).

To select the correct statistical tests to run on the data in your dissertation, we have created a Statistical Test Selector to help guide you through the various options.

Prepare and analyse your data using a relevant statistics package

The preparation and analysis of your data is actually a much more **practical** step than many students realise. Most of the time required to get the **results** that you will present in your **write up** (i.e., usually in Chapter Four: Results) comes from knowing **(a)** how to **enter data** into a statistics package (e.g., SPSS) so that it can be analysed correctly, and **(b)** what buttons to press in the statistics package to correctly **run** the statistical tests you need:

Entering data

Entering data is not just about knowing what buttons to press, but:

**(a)**how to**code**your data correctly to recognise the**types of variables**that you have, as well as issues such as**reverse coding**;**(b)**how to**filter**your dataset to take into account**missing data**and**outliers**;**(c)**how to**split files**(i.e., in SPSS) when analysing the data for separate subgroups (e.g., males and females) using the same statistical tests;**(d)**how to**weight**and**unweight**data you have collected; and**(e)**other things you need to consider when entering data.

What you have to do when it comes to entering data (i.e., in terms of coding, filtering, splitting files, and weighting/unweighting data) will depend on the statistical tests you plan to run. Therefore, entering data starts with using the Statistical Test Selector to help guide you through the various options. In the Data Analysis section, we help you to understand what you need to know about entering data in the context of your dissertation.Running statistical tests

Statistics packages do the hard work of statistically analysing your data, but they rely on you making a number of choices. This is not simply about selecting the correct statistical test, but knowing, when you have selected a given test to run on your data, what buttons to press to:

**(a)**test for the**assumptions**underlying the statistical test;**(b)**test whether**corrections**can be made when assumptions are**violated**;**(c)**take into account**outliers**and**missing data**;**(d)**choose between the different**numerical**and**graphical**ways to approach your analysis; and**(e)**other standard and more advanced tips.

In the Data Analysis section, we explain what these considerations are (i.e., assumptions, corrections, outliers and missing data, numerical and graphical analysis) so that you can apply them to your own dissertation. We also provide**comprehensive**,**step-by-step instructions with screenshots**that show you how to**enter data**and**run**a wide range of statistical tests using the statistics package, SPSS. We do this on the basis that you probably have little or no knowledge of SPSS.

Interpret the findings properly

SPSS produces many tables of output for the typical tests you will run. In addition, SPSS has many new methods of presenting data using its Model viewer. You need to know which of these tables is important for your analysis and what the different figures/numbers mean. Interpreting these findings properly and communicating your results is one of the most important aspects of your dissertation. In the Data Analysis section, we show you how to understand these tables of output, what part of this output you need to look at, and how to write up the **results** in an appropriate format (i.e., so that you can answer you research hypotheses).