User research: Comparison of different groups completing questionnaires in various periods

Article Description: probe into the time of the research questionnaire.

Written in front: This study is not involved in the weeks of more reasonable, there are two main points to consider: 1, the day-to-day research projects have the basic implementation cycle of the limit, generally are the more able to launch as soon as possible, and the basic work in the working day, 2, the current implementation of the research projects, weeks, if the time allowed, will continue to collect data for a week, The overall opening rate fluctuates little.

Introduction

In the daily research work, in order to improve the scientific nature of the questionnaire, increase the randomness of users of questionnaires, for the buyer's questionnaire survey, we will generally use email to send research invitations; for sellers, they will use the way of station letter and email. At the same time, in order to reduce the harassment of users, they can opt to unsubscribe from such information.

We will have a simple judgment, that is, the peak time users surf the internet, the peak time to browse Taobao, but also they open the survey of the peak time, but the actual situation is this?

In this study, we selected 5 buyers ' research, 4 sellers ' research, and conducted a thorough study on the time for users to open the questionnaire. Due to the different time points of these 9 surveys, the data should be processed before the study and the samples in the complete cycle (24 hours *n days) should be retained.

The time distribution of the research questionnaire opened by Taobao users

Taobao's traffic peaks typically appear at midday breaks, before bedtime, and, at the same time, sellers ' online peaks are comparable to those of buyers.

A recent Taobao buyer survey, users fill out their own online time period, browse Taobao time period as follows (focus on the trend can):

You can see that the survey sample and Taobao overall user browsing time period of Taobao, there is no significant fluctuation trend , but also shows that the randomness of the research sample is very good, can represent the overall user. Can come to the conclusion: Taobao users surf the time period and browse Taobao time period, the trend of the wave basically coincide.

and summed up 5 buyers research, 4 sellers research data, you can find that Taobao users to open a survey of the peak period is not the Internet/browse Taobao peak period .

In general, Taobao buyers Open the survey of the time period, mainly concentrated in 8-10 points, followed by 10-12 points, in addition, 12-14 points, 14-16 points, 16-18 points, 20-22 points open rate is also high.

Taobao sellers open a survey of the time period, mainly concentrated in 8-10 points, 10-12 points, followed by 12-14 points, 14-16 points.

Taobao users in each of the full time to fill in the survey questionnaire comparison

To investigate the results of questionnaires, in addition to the opening rate, you also need to see the full answer rate, that is, the complete sample size/Open Sample Total *100%, if a certain period of time the full filling rate is very low, even at this time the opening rate is higher, is not an ideal state.

After analysis of variance 22, we know that there is no significant difference in the total filling answer rate in the main time period of each investigation. as shown in figure:

Taobao buyers in the main filling period, the full answer rate of nearly 80% average, Taobao sellers in the main fill in the answer period, the average of the full answer rate of about 75%.

The comparison of the length of time when the survey questionnaire was completed by Taobao users

Completion of the survey of the length of the questionnaire to affect the quality of the questionnaire, if the completion of a period of time to complete the questionnaire is too short or too long, even if the opening rate is higher, the full answer rate again high, nor is the ideal state.

The results of variance analysis 22 showed that there was no statistically significant difference in the time of completion of each survey.

The length of each survey is different, the time required to complete the questionnaire is different, so the average completion time of each survey is meaningless.

Comparison of questionnaires completed by different groups in each period

Some questionnaire research focus on the target population is special groups, such as low-star buyers, full-time sellers and so on, then different types of groups, in each period to complete the proportion of the questionnaire is there any difference? (Group A in a certain period of time to complete the questionnaire ratio = Group A in a certain period of time to complete the questionnaire sample/Group A completed Questionnaires total sample)

By variance Analysis 22 test, comprehensive analysis, the buyer survey, in the different identities of Taobao (pure buyers, have bought mainly sellers, both buy and sell sellers) between groups, different buyers star groups, in the main filling out the proportion of the questionnaire, basically there is no significant difference.

Seller research, in the different identities of Taobao (pure sellers, both sell and buy sellers, have bought the main seller) between groups, different sellers star groups, different formal investment Taobao time between the group, in the main filling out the proportion of the questionnaire, basically there is no significant difference.

Summary:

1, Taobao users to open a survey of the peak period at 8-10, 10-12 points, and the main period of the full answer rate, the completion of the length of no significant differences. Future research questionnaires, you can change the current no exact delivery time, one-time delivery of the situation, will be put into a number of samples, batch time every morning, sent to complete just to catch the user's peak period, to better fill the answer effect. emergency projects are particularly applicable.

2, in accordance with the specific identity of Taobao, or buyers (sellers) star, the official Taobao shop when the length of the factors such as buyers, sellers groups for research, that is, the study of special groups, the time to put the questionnaire without special arrangements, and the study of the entire network of users of the same period of

Note:

1 , the conclusion of this study is only suitable for similar shopping sites such as Taobao, for other types of Web sites, there may be a large difference, you can use the method mentioned in this paper for another study.

2 , there are many factors that affect the opening rate of a certain questionnaire, such as the accuracy of the target group, the relevance of the subject and the user's vital interests, whether there is incentive for prizes, the description of the title and content of the invitation, whether the questionnaire is too long, etc.

In this paper, the questionnaire survey used to synthesize a variety of situations, and strive to find a normal, so the subsequent application of the conclusions of this article, even in the case of special cases of questionnaire research, this normal will still work.

3. In addition, it is undeniable that no matter which day, which time to put the questionnaire, the beginning stage, the rate of open questionnaires are the highest, but this is based on a large number of the base to do support, and did not achieve the effect of saving resources, the application of the conclusions of this article, not only to use the advantages of the beginning stage, but also to save resources