Tips for when encoding self-administered survey data becomes less straightforward

Encoding simple surveys, such as quick activity evaluations, done on paper into electronic format is a fairly straightforward task – just read the numbers or text and plug them into a spreadsheet. It of course does require a great deal of concentration (so you don’t get lost and start encoding the wrong rows, which completely messes up any subsequent analysis) and some intuition (when trying to decipher bad handwriting).

But for the benefit of the people who are subsequently going to look at the data, please do the following as well:

1. If a response is blank, just leave it blank.

2. Have a standard thing to write for responses such as “n/a”, “na”, “none”, “nothing”, “-“, etc. These should all be encoded as the same thing – for example, just “NA”. Unless you’re specifically studying different kinds of ways of writing “none”, or have reason to believe that “N/A” and “nothing” have a significant qualitative difference, having a single code for these types of responses should be sufficient. Important especially if more than one person is encoding.

3. When encoding numerical data, such as surveys that ask you to rate something from a scale of 1 to 5, some respondents break the rules and do things like write the number 6 and encircle that, or encircle 3 and 4 at once to denote a “3.5”. If you see responses like these, do not encode them. The safest thing to do is to just discard them by classifying them as blank, because you cannot really guess what the respondent was thinking. Or at your discretion, you might classify a “6” as a “5”, or a “0” as a “1”. Just please don’t actually encode “6”, or “3 and 4”.

If you’re an encoder and you encounter an odd situation that you’re not sure how to handle, ask your supervisor first – but make sure that whatever your supervisor recommends is then shared across everybody else encoding. Consistency is key.

4. When encoding textual data, leave the text exactly as is, including slang, acronyms such as OMG, odd punctuation, and typographical errors. Yes, please leave typos exactly as is – resist the urge to correct a misspelled word. Sometimes the presence of typographical errors can be an important point in analysis.

All of these tips are simple to keep in mind and will reduce ambiguity in encoding, as well as lessen the amount of work needed to clean up the data afterwards for analysis.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s