Step 2: Identify and Group Respondents
Having established a goal for the study, it is time to identify the right people to include in the survey. Here it is important to bear in mind the scope of the study and that the complexity of the network rises as more people are included.
If the goal of the study is very specific (e.g. to improve cross sales) it should not be too difficult to identify the relevant employees, if the study is more generic the number of respondents can be limited by focusing on specific departments or organizational levels.
In general, social network visualizations with fewer than 25 nodes are easy to decipher and statistics are not necessary. With more employees than that, the statistics become increasingly helpful for understanding the network; at 100 employees you will rely almost solely on statistics. A study with more than 150 nodes requires the help of an SNA consultant.
If the study involves more than 50 respondents it can be a very good idea to limit the number of rating questions to two or three. This is because it becomes more time-consuming for the respondent to answer the questionnaire as more people are added to the study.
How you group respondents is important. The groups to which you have assigned employees are used both in the network statistics and the visualization. Groups determine the color of nodes in the social network diagram and in the statistics you can see how these groups interact on an overall level. Most commonly, departments are used as the way to group people, but there is no limitation on how to group respondents. Other ways to group employees include: organizational level, location and years of employment. It is possible to collect data for grouping respondents with the survey too so the respondents themselves deliver this information.
On the People page in Socilyzer a group is first created and then employees are added to it. Employees are added with name and email either manually (one at a time) or by uploading a spreadsheet. The email is used when the questionnaire is sent out and the respondent’s name is displayed in the questionnaire for the others to rate their interaction with.