Resultal tutorial: How to get the most out of your data

So you formulated your research question, created the experiment and collected your data. What happens next?

Even though Resultal offers real-time dashboard visualizations, you might want to do some additional analysis on your data. In that case you need to download the raw data which is directly formatted as a comma-separated text-file (a.k.a CSV-files). In this short blog, I will go over the standardized format created in Resultal and explain how you can get the most out of your data!

To download your data go to the Data / Downloads icon in the menu on the left (arrow 1) and click on ‘Download experiment as csv’ (arrow 2). You can select the data from the entire experiment, or select one of the individual tasks. Since we don’t have the same data, let’s consider the following set. This data was collected for a Bachelor-thesis study on the topic of sustainable consumerism.

To get your raw data, first click on the icon in the menu on the left, then select ‘Download experiment as csv’.

As soon as you found your data and opened it within Excel (or any similar spreadsheet software), the standardized CSV-file might look quite confusing. This is because all the data is in the first column, which makes it difficult to read. To make more sense out of it, you need to create additional columns for each of the different variables. Luckily, Excel has a very simple tool for that: Text to columns

• Click on the first column (‘A’)
• Go to the Data tab
• Find and select the ‘Text to columns’ function
• In the menu select ‘Delimited’ and click next
• In the delimiters menu, tick ‘Comma’ and click next
• In the next menu, click ‘finish’ and see the magic happening

The raw data as you would download it, which is difficult to read.

The raw data after ‘text to columns’, much better to see!

The variables and values

Now we have a clearer picture on what is actually in the data. The first row of the file contains the variable names. Let’s go over each of the columns step by step to address the type of variable, what the values represent and what you can use it for in your analysis. At first glance it might seem overwhelming but as soon as you are familiar with the files it makes a lot more sense:

• id. The first variable represents the number of your data row. It is sequential and mostly used to retrace every single line that has been written away in our database. Ergo, it is the least important for you.
• category. This variable contains the category of the question that was presented to the participant. If it is empty it means that there was no category involved. Currently, there is only one Task type that uses categories (‘Statements with categories’)
• user_id. Your unique researcher number.
• visitor_id. The unique participant number.
• response. The response provided by the participant. This value depends on the Task type (i.e. what you asked the participant).
• attribute. This is the characteristic associated with the response of the participant. Currently, there is only one type of attribute: the location of the response (left/right. It is only used if the answer options were two statements (buttons), otherwise the variable remains empty.
• response_time. Reaction time of the participant. The counter starts at the moment the question is presented and is saved when the participant gives a response. It is measured in milliseconds (ms). (Please note: In a previous build this variable was called ‘duration’)
• task_id. Your experiment’s unique task identifier (Please note: In a previous build this variable was called ‘module_id’)
• experiment_id. The unique identifier of your experiment.
• time_elapsed. Time elapsed since the start of the experiment. Measured in milliseconds (ms).
• created_at. The date and time the participant finished the experiment. It is accurate up until the minute, it is finished. Note that your data is only written and saved after your participant finished the experiment. This way we can assure that both you and your participants have full control and access to the data.
• wid. The unique WorldBrainWave identifier, use it to easily track or filter specific tasks or even specific questions. The identifier is composed as experiment_id + task_id + question_id. Every wid in the database is unique, as a result you can always trace back and access your data.

As you can see, there is quite a lot of data included in the files, but once you understand the function of each variable it makes analyses much simpler. In most cases you only need the following essential variables for your analysis:

1. wid (“What did you ask?”)
2. response (“What did they respond?”)
3. response_time (“How long did it take them?”)

As soon as you can filter and understand these three variables, you can start your analysis in Excel, SPSS, JASP, R or any other (statistical) software. As you can see the CSV-file is always written in such a way that every question or trial is saved on one single line of data, this is the so-called long data format. In the next post I will discuss pivoting the data into a wide data format. If you conduct your (statistical) analysis in SPSS it is essential to pivot your data.

TL:DR

All raw data files generated by Resultal or WorldBrainWave are in a pre-constructed format and written in CSV, seperated with commas. To download your data go to the Data / Downloads menu and find your experiment. This file needs to be divided into readable columns in Excel with the text-to-columns function. Finally, there are three essential variables for your analysis: wid, response and duration. The first value gives the unique identifier of your question, the second the response of the participant and the third how long took them to respond.

You can start right away with your first experiment at Resultal. If you have more questions or feedback, feel free to contact us through Twitter or LinkedIn