Statistical Questions and Tidy Data

We have some data obtain from the resource available here.

This shows the language diversity within different countries as well as their population and area.

In the last section we learned about tidy data and how it is determined based on what we are trying to figure out, eg: what our statistical question is.

There are cases where a longer dataframe is considered tidy data and in other cases, a wider dataframe is tidy depending on what is being ask.

Let’s explore some more examples.

We are trying to answer the statistical question:

“What factors are associated with language diversity (as measured by the number of languages spoken in a country)?”

We are trying to answer the statistical question:

“What factors are associated with language diversity (as measured by the number of languages spoken in a country)?”

In this case, it makes the most sense for our data to have the measurements area and population as variables that may or may not affect language diversity. This would be considered tidy data and it is stored as a wider dataframe than what was initially presented to us.

UNICEF dataset on under-five child mortality available here.

Let’s look at another example.

Here is the UNICEF dataset on under-five child mortality which is available here. It is based on data that can be found at www.childmortality.org. The under-five mortality rate is expressed as the number of under-five deaths per 1,000 live births.

The statistical question that we’d like to answer with this data set is:

“Is there a relationship between country and/or year of under-five child mortality?”

The statistical question that we’d like to answer with this data set is:

“Is there a relationship between country and/or year of under-five child mortality?”.

We are interested in the variable year. Having a single variable for year would be the most appropriate to use in this analysis and display tidy data. This would correspond to a long dataframe.

At this point you may be asking yourself the question “Are there any verbs, that would easily transform data into longer or wider dataframes?”

If you are, then you are ahead of the game! We will discuss this in the next section.

Let’s apply what we learned!