You’ve collected hundreds of raw open-ended replies to your survey, or ideas submitted to your suggestion box, or feedback from your customer satisfaction follow-up. Now what? To make sense of all of it you need to categorize and classify the replies, allowing you to determine the most common and salient opinions, meanings and topics that were shared. Tabulation of distilled verbatim text can also lead to actionable insights that may not be drawn from a casual read-through. The response coding service offered by Princeton National Surveys handles this task when you want to analyze . . .
- Replies to questions that you’ve conducted on online survey platforms
- Employee or customer feedback that you’ve collected from online suggestion box apps
- What your audience or customers are saying in reaction to your content on social media platforms
- Online reviews and ratings of your product or service
- Thoughts and feelings written on comment cards you scanned and saved as image files
How We Help
Princeton National Surveys’ coding service turns the unrefined richness of free-text replies accumulated from any feedback platform into clean categorical data for quantitative analysis. We use the industry standard practice of semantic coding of natural language text, not artificial processing. A squad of trained and experienced coders, vetted and reviewed by specialists with 20 years of experience in survey research, works on your project. The advantages of this approach:
- Coders know what was asked and why, or what experience or circumstance from which the comments were elicited.
- Coders focus on and extract the most important nuggets, ignoring the background noise and insignificant parts of speech.
- Coders can recognize if a reply refers to a name (person, brand, company) rather than a common word, even if it is misspelled.
- Coders can cope with poorly written answers with spelling and grammatical errors, acronyms (LOL, WTF) and the insertion of emoticons/emoji by respondents.
- Coders can detect sarcasm in what would appear to be expression of a positive (or negative) sentiment.
- Coders can flag gibberish, nonsense and non-responsive entries.
How We Work
Our coders distill the raw content into specific points, themes or generalized concepts based on a classification framework – one you provide or one that we develop with your approval. If you provide a complete list of primary ideas, elements and details to extract from the verbatim replies, work can begin within an hour of receipt (during regular business hours). Or, we can develop a list based on a preliminary read-through of the replies, modify this list with your review and input, and get the work started within an hour after final approval.
Our response coding service is of particular value when your project entails:
After coders complete their work and the data are reviewed, we will send you the coded responses in Excel. You may choose to receive the data in columns of 1s & 0s, or columns of Codelabels & 0s. A codebook with frequencies and percentages is included as well. The order of the codes will be arranged in decreasing frequency (high to low). SPSS and Triple-S formats can also be prepared.
The cost and amount of time to process and deliver data back to you will depend on a few factors: the number of unique text answers, the number of different questions that were asked, whether or not you provide a coding framework, and the length and the complexity of the raw verbatims to be coded. Turnaround time for the first 1000 (or fewer) text answers is three business days; add one more business day for 1001-2000, a fourth business day for 2001-3000, and so on. Some additional time is required if we develop the coding strategy for you.
Please fill out and submit the form below so we may more accurately estimate how quickly we can deliver results to you and at what price.