Self-learning surveys for open-ended analysis

1. A computer-implemented method of automatically evolving survey questions, comprising:

receiving, at a processor, a first set of textual responses to at least one open-ended question in a survey;

reformatting the first set of textual responses as a set of possible textual responses in the form of closed-ended questions with numerical ratings as options;

storing the set of possible textual responses in a memory in communication with the processor;

selecting and transmitting a subset of the set of possible textual response to survey recipients;

receiving, at the processor, participant numerical ratings for the possible textual response;

using the processor to generate a confidence measure for each possible textual response based upon the participant numerical ratings;

generating a ranked list using the confidence measure, with the processor;

storing the ranked list of textual responses; and

converting at least one of the original open-ended questions in the initial version of the survey to a closed-ended question, using a subset of the textual responses provided.

2. The computer-implemented method of claim 1 , wherein selecting and transmitting a subset of possible textual responses comprises selecting and transmitting multiple subsets of different possible questions to different participants.

3. The computer-implemented method of claim 1 , wherein receiving the participant ratings for each of the possible textual responses further comprises computing a confidence measure of the ratings.

4. The computer-implemented method of claim 1 , wherein the confidence measure is based on a standard error of the mean.

5. The computer-implemented method of claim 1 , wherein generating the ranked list comprises:

gathering a set of possible textual responses to a question;

determining a standard error of the mean for each of the possible textual responses; and

prioritizing presentation of low confidence textual responses in subsequent versions of the survey.

6. The computer-implemented method of claim 1 , wherein converting the at least one open-ended question to the closed-ended question comprises selecting a subset of textual responses that have a highest value of mean numerical rating values.

7. The computer-implemented method of claim 6 , wherein converting the at least one open-ended question to the closed-ended question comprises selecting a subset of textual responses that have a highest Wilson score.

8. The computer-implemented method of claim 1 , wherein the receiving and reformatting continues during the selecting and transmitting the subset of possible textual responses.

9. A computer-implemented method of automatically evolving survey questions, comprising:

receiving, at a processor, a first set of possible questions for a survey;

selecting and transmitting a subset of the first set of possible questions to survey recipients;

receiving, at the processor, participant ratings for each of the possible questions;

using the processor to generate a confidence measure for each question based upon the participant ratings;

generating a ranked list, with the processor, using the confidence measures;

storing the ranked list; and

selecting a predetermined number of the ranked question list and presenting those on a next version of the survey.

10. The computer-implemented method of claim 9 , wherein selecting and transmitting a subset of possible questions comprises selecting and transmitting multiple subsets of different possible questions to different participants.

11. The computer-implemented method of claim 9 , wherein receiving the participant ratings for each of the possible questions further comprises computing a confidence measure of the ratings.

12. The computer-implemented method of claim 9 , wherein the confidence measure is based on a standard error of the mean.

13. The computer-implemented method of claim 9 , wherein generating the ranked list comprises:

gathering a set of possible question for a survey;

determining a standard error of the mean for each of the possible questions; and

prioritizing presentation of low confidence responses in subsequent versions of the survey.

14. An automated survey system, comprising:

a first computing system having a processor, the first computing system transmitting an access to an electronic survey having at least one open-ended question;

at least one second computing system to receive the access to the electronic survey and to access the survey;

the processor to execute instructions causing the processor to:

receive a first set of textual responses to the at least one open-ended question;

reformat the first set of textual responses as a set of possible textual responses in the form of closed-ended questions with numerical ratings as options;

store the set of possible textual responses in the memory in communication with the processor;

select and transmit a subset of the set of possible textual response to survey recipients;

receive, at the processor, participant numerical ratings for each of the possible textual responses;

use the processor to generate a confidence measure for each possible textual response based upon the participant numerical ratings;

generate a ranked list, with the processor, including the confidence measures;

store the ranked list; and

convert at least one of the original open-ended question in the survey to a closed-ended question, using a subset of the textual responses.