LatanaLatana

Data collection and methodology

We regularly update and add features to the Latana dashboard, and we collect all learning materials in this organized database.

Data collection and methodology

How does Latana process open-ended questions?

Firstly, we design open-ended questions in a way that minimises nonsense answers in the data. There is an instruction to respondents included next to the question that advises respondents to answer “no” if they cannot think of any answers to the question provided to ensure we can correctly differentiate them from nonsense answers.

Once all responses are collected, we run a script to check for any nonsense answers we do get and clean anything out that is not a proper answer/looks like nonsense. We then sense-check the results following this cleaning, to ensure that no valid responses have been inadvertently cleaned and that no nonsense answers still remain in the clean data.

Lastly, we run a proprietary script that correctly categorises each answer allowing for some different spellings and misspellings before the answers are translated via our MRP process into the results visible in our App.