Using Raw Data in Research: Importance and Examples

Why Raw Data is Important in Research

When you are looking for your data, it is important that you are using the raw data and not the analysed data. If the data you are looking at is already reported as averages or percentages, this is not raw data. Typically, when you are looking for data, you will find the analysed data first (aggregate counts, averages, percentages), but for some organisations, you are able to download the raw data that was used to produce the analysis.

The raw data will have individual data for every participant or unit in the data. For example, I have included, as an SPSS file, in the Assessment 1 Brief folder the raw data for the Air Passenger Survey 2019, which is data that is relevant for both aviation and tourism students. There is individual data for over 44,000 participants for this survey, which can be seen in the SPSS ‘Data View’ tab. There are also 13 variables, which can be seen in the ‘Variable View’ tab.

You are welcome to use this data for your Insights Reports, or you can find your own data. Just make sure that you have found and are using raw data; not data that has already been analysed. Although you could use analysed data for Assessment 1 because you will be able to report descriptive statistics, you would not be able to run the inferential statistics required for Assessment 2, which requires analysis of the participant or unit level data that is only available for raw data.

You are permitted to use the same client, data and analysis for Assessment 2 as you do for Assessment 1. In other words, you will not be held accountable for any self-plagiarism. However, if you use analysed data (and not raw data) for Assessment 1, you will lose this opportunity.

As an example, the Air Passenger Survey 2019 using the following variables, which are either categorical variables or continuous numerical variables:

Continuous Numerical Variables

·Air Miles

·Length of Stay

·Spend on Visit

There are many questions that can be asked using this data.

Count Research Questions (Descriptive Statistics)

You could do counts on the categorical variables. For example:

What are the destinations (dependent variable) of airline passengers from the UK?

This tells us the count and percentage of passengers for the destinations.

This could also be done for all of the categorical variables, but you have to think about the practical importance of asking such questions.

Examples of Raw Data from Air Passenger Survey 2019

Cross-Sectional Research Questions

You could also look at the influence of categorical variables on continuous numerical variables. For example:

What is the influence of destination (independent variable; top 10 destinations) on length of stay (dependent variable) for UK air passengers (participants & setting)?

1.Analyze > Compare Means > Means

2.Dependent List: Length of stay

3.Independent List: Main Country visited

Such analysis could be conducted by pairing any categorical variable with any continuous numerical variable, but again, make sure the answer to any question is practically important.

Client focus

You are required to choose a client for your Insights report, so it may be that you want to restrict the data that you are using. For example, you may decide that your client is the Spanish Tourist Board and you want to only look at data on UK air passengers to Spain and their direct competitors (e.g., Canary Islands and Portugal).

For example, you might want to start by answering the following count research question:

What are the destinations (dependent variable; Spain airports) of UK tourists (participants) to Spain (setting)?

For ease of analysis (i.e., not having to scroll through all the data) you can restrict your data to just participants who identified Spain as their destination, as follows:

Then, you can run your analysis based purely upon UK tourists to Spain:

Next, you might be interested in a comparison with Spain’s primary competition of the Canary Islands and Portugal. For example:

What is the influence of destination (independent variable; Spain, Canary Islands, Portugal) on spend (dependent variable) for UK air passengers (participants & setting)?

There is no need to restrict your data as it is simple enough to just look up the mean spend for the destinations you are interested in, so go back into your data and release your restriction:

What is the influence of destination choice (Lugano, Como or Milan) on age of tourists to Southern Switzerland/Northern Italy?

This question represents a cross-sectional research design looking at the influence of a categorical variable (destination choice) on a discrete variable (age recorded to nearest year) with both variables recorded at a single point in time.

To answer this question, using statistical software, the ages of respondents who were visiting each of the destinations of interest (Lugano, Como, Milan) were summed separately for each destination choice and then divided by the number of respondents for each destination choice to return a Mean average of age by destination choice (Buglear, 2019). The statistical software also calculated the Standard Deviation for the Mean. Comparisons were made and percentage differences between Means were calculated.

What is the influence of age category on importance of core destination choice motivations (relaxation, natural beauty, outdoor activities, cultural, social) for tourists to Southern Switzerland/Northern Italy?

This question represents a cross-sectional research design looking at the influence of a categorical variable (age group) on a discrete variable (Likert scale for importance of destination choice motivation) with both variables recorded at a single point in time.

To answer this question, using statistical software, the strength of each destination choice motivation (relaxation, beauty, activities, cultural, social) were summed separately for each age group (18-30, 31-50, 51-70, 71+) and divided by the number of respondents for each age category to return a Mean average of strength of each destination choice motivation by age group (Buglear, 2019). The statistical software also calculated the Standard Deviation for the Mean. Comparisons were made and percentage differences between Means were calculated.

What is the influence of destination (Lugano, Como or Milan) on strength of attributes relaxation, natural beauty, outdoor activities, cultural, social) for tourists to Southern Switzerland/Northern Italy?

This question represents a cross-sectional research design looking at the influence of a categorical variable (destination choice) on a discrete variable (Likert scale for strength of destination choice motivation) with both variables recorded at a single point in time.