Webinar Wire

Features: Polling

Posted by Ken Molay December 12 2007 8:40 PM EDT

Interactive audience polling can give you extra information about your viewers as well as increasing their level of participation and concentration on your material. Today I will examine differences in implementation approaches for polling among web conferencing vendors.

The first distinction is what types of questions you can ask (or more accurately, what types of answers you can receive). The most common implementation of web conference polling allows audience members to select one answer from a multiple choice list. Some vendors only allow you to create a poll from a preconfigured answer template. In this case, your audience must pick one of the set answers in the provided question template. These might be choices such as YES/NO, TRUE/FALSE, or HIGH/MEDIUM/LOW. You cannot customize the response labels or the number of choices available.

[A very few vendors don't even let you display a question and answers to the audience. They ask you to announce your question vocally, with the audience setting a flag such as a "raised hand" icon to signal their response. This type of polling is analogous to surveying raised hands in a live conference room and barely fits into the concept of interactive polling as a supported conferencing feature.]

Adding a bit more flexibility, some vendors let you create your own list of answer labels, but limit you to a preset maximum number of answer choices (typically five to seven). With more flexibility still, some vendors let you add as many answer choices as you want, with no limit.

I am a big fan of web conferencing packages that let the presenter specify a multiple choice question as single-answer or multi-answer. There are many cases where you will want to ask the audience to "pick all that apply".

There are various approaches to closing the voting period for polls. Some vendors apply a fixed number of seconds that the poll remains open for voting. Some let the presenter manually close the voting period so that no more votes can be entered. Some leave out the notion of closing a poll. This implementation feature may be important to you if you are conducting true voting, such as in an election, where you need to announce a formal resolution based on the results.

You will also see implementation differences in whether the audience can change their votes once submitted. Some products allow them to switch their votes as long as the poll is open and some force them to submit an entry, which is unchangeable thereafter. I have a personal preference for allowing people to change their minds or correct a mis-key, but that's my own bias.

I also like polling implementations that show the presenter how many people out of the total audience have submitted a vote. As the number changes in real-time, you can estimate when it is reasonable to comment on the results or move on.

Most web conferencing products let presenters see the results without displaying them to the audience. This should be a requirement on your checklist. If you can't see the audience answers until you show them to everybody, you can't ask questions that could imply negative commentary about you or your material. It's a good idea to check in during a long presentation and ask how the pace is, but not if the audience can see the fact that everybody thinks you are dragging terribly!

The manner in which results are displayed is another variant in conferencing products. Some only show the absolute number of votes associated with each answer. Others only show percentage figures next to each answer. Many products display a bar graph, and TalkPoint lets you choose whether to show percentage results in a bar graph or pie chart. For structured webinars, my preference is to have the option of whether to show absolute numbers and/or percentages. But if only one of the two is available, it should definitely be the percentage results. You don't want your audience to know that you had a terrible turnout and there are only eight people watching. It's much better to be able to say that "out of all of you, 28% thought..."

As with every other feature in the user interface, the question of where to show the poll and results comes up. The traditional method originally used by Placeware and WebEx was to create the poll as a special slide in your PowerPoint presentation. The contents were converted to the product's interactive format during the slide upload process and the poll would be encountered in sequence as you moved through your slide deck. Most companies have gone to an approach that keeps polling questions separate from the slide presentation content. They may appear in the framing portion of the interface window or each can pop up in its own mini-window. The Adobe Connect philosophy of infinitely customizable size and placement of "pods" and "layouts" in the user interface means that you can place polls almost anywhere that looks good to you as an organizer.

You should also test the creation process for adding a new polling question. Ideally you want the flexibility to create and store polling questions before an event starts so you can select the desired one during the event, as well as being able to create a new poll on the fly during an event (just in case the need arises). Vendors may allow one or the other method, or they may allow both options.

The last thing to investigate when checking your vendor's polling implementation strategy is how results can be reported after your event ends. Ideally you should be able to download nicely formatted reports that show summary results for each question, as well as the individual responses by attendees. I appreciate it when these can be extracted to Excel for further analysis, sorting, and grouping.

I'm keeping the issue of standalone polling questions separate from more extensive surveys and quizzes. We'll save that for another features column.