A data analysis is an evaluation of formal data to gain knowledge for the bachelor’s, master’s or doctoral thesis. The aim is to identify patterns in the data, i.e. regularities, irregularities or at least anomalies.
Data can come in many forms, from numbers to the extensive descriptions of objects. As a rule, this data is always in numerical form such as time series or numerical sequences or statistics of all kinds. However, statistics are already processed data.
Data analysis requires some creativity because the solution is usually not obvious. After all, no one has conducted an analysis like this before, or at least you haven't found anything about it in the literature.
The results of a data analysis are answers to initial questions and detailed questions. The answers are numbers and graphics and the interpretation of these numbers and graphics.
What are the advantages of data analysis compared to other methods?
- Numbers are universal
- The data is tangible.
- There are algorithms for calculations and it is easier than a text evaluation.
- The addressees quickly understand the results.
- You can really do magic and impress the addressees.
- It’s easier to visualize the results.
What are the disadvantages of data analysis?
- Garbage in, garbage out. If the quality of the data is poor, it’s impossible to obtain reliable results.
- The dependency in data retrieval can be quite annoying. Here are some tips for attracting participants for a survey.
- You have to know or learn methods or find someone who can help you.
- Mistakes can be devastating.
- Missing substance can be detected quickly.
- Pictures say more than a thousand words. Therefore, if you can’t fill the pages with words, at least throw in graphics. However, usually only the words count.
Under what conditions can or should I conduct a data analysis?
- If I have to.
- You must be able to get the right data.
- If I can perform the calculations myself or at least understand, explain and repeat the calculated evaluations of others.
- You want a clear personal contribution right from the start.
How do I create the evaluation design for the data analysis?
The most important thing is to ask the right questions, enough questions and also clearly formulated questions. Here are some techniques for asking the right questions:
Good formulation: What is the relationship between Alpha and Beta?
Poor formulation: How are Alpha and Beta related?
Now it’s time for the methods for the calculation. There are dozens of statistical methods, but as always, most calculations can be done with only a handful of statistical methods.
- Which detailed questions can be formulated as the research question?
- What data is available? In what format? How is the data prepared?
- Which key figures allow statements?
- What methods are available to calculate such indicators? Do my details match? By type (scales), by size (number of records).
- Do I not need to have a lot of data for a data analysis?
It depends on the media, the questions and the methods I want to use.
A fixed rule is that I need at least 30 data sets for a statistical analysis in order to be able to make representative statements about the population. So statistically it doesn't matter if I have 30 or 30 million records. That's why statistics were invented...
What mistakes do I need to watch out for?
- Don't do the analysis at the last minute.
- Formulate questions and hypotheses for evaluation BEFORE data collection!
- Stay persistent, keep going.
- Leave the results for a while then revise them.
- You have to combine theory and the state of research with your results.
- You must have the time under control
Which tools can I use?
You can use programs of all kinds for calculations. But asking questions is your most powerful aide.
Who can legally help me with a data analysis?
The great intellectual challenge is to develop the research design, to obtain the data and to interpret the results in the end.
Am I allowed to let others perform the calculations?
That's a thing. In the end, every program is useful. If someone else is operating a program, then they can simply be seen as an extension of the program. But this is a comfortable view... Of course, it’s better if you do your own calculations.
A good compromise is to find some help, do a practical calculation then follow the calculation steps meticulously so next time you can do the math yourself. Basically, this functions as a permitted training. One can then justify each step of the calculation in the defense.
What's the best place to start?
Clearly with the detailed questions and hypotheses. These two orient the entire data analysis. So formulate as many detailed questions as necessary to answer your main question or research question. You can find detailed instructions and examples for the formulation of these so-called detailed questions in the Thesis Guide.
How does the Aristolo Guide help with data evaluation for the bachelor’s or master’s thesis or dissertation?
The Thesis Guide or Dissertation Guide has instructions for data collection, data preparation, data analysis and interpretation. The guide can also teach you how to formulate questions and answer them with data to create your own experiment. We also have many templates for questionnaires and analyses of all kinds.
Good luck writing your text!
Silvio and the Team Aristolo
PS: Check out the Thesis-ABC and the Thesis Guide for writing a bachelor or master thesis in 31 days.