Months ago, I mentioned that I was part of an ITS pilot of PollEverywhere this past spring. Quick reminder: PollEverywhere is a web-based service where anyone can create a multiple-choice or open-ended question and people can respond via text, Twitter or website. I first used PollEverywhere in the fall when I wanted a way for my teams to submit open-ended responses. The free version only allows up to 30 responses per poll which was fine for 13 team responses but wouldn't work for individual responses (since I have 75 students in each of my sections) so I used clickers for any individual responses. In the spring, the University bought a PE account subscription so there could be unlimited responses. It also meant that students could register and their responses were recorded so I could use PollEverywhere as a replacement for clickers. In this post, I'll explain the mechanics of how I used PollEverywhere and some of the associated pluses and minuses. In my next couple posts, I'll talk about what the students thought and my overall impressions.
Low-stakes assessments: I mostly used PE to have students submit their individual responses to multiple-choice application problems before they discussed those same problems in their teams. My main concern was making sure that students had to think about the problem individually a little and commit to an answer before discussion. I didn't care so much what specific answer they chose so students received credit just for answering anything (i.e., participating); there were only a couple times when I made credit dependent on selecting the 'right' answer. PE allows you to embed polls in PowerPoint and that is what I did, rather than switching over to the website each time. One downside of PE, relative to clickers, is that there is no timer so I created one using animation in PowerPoint. It's a clunky workaround but if you want to give students a visual indication of how much time they have to answer a question, I'm not sure what the alternatives are.
Grouping questions: One thing I had to decide was if I wanted to 'group' my questions together or not. The way PE normally works, each answer choice has a randomly-generated unique keyword; for example, if you want to choose answer A, then you send in '70101' and if you want to choose answer B, then you send in '70103', etc. With a paid account, you can also create your own keywords to replace the random numbers (but they still have to be unique since the keyword identifies both the question and the specific answer choice). An alternative is grouping multiple questions together and assigning a keyword to the group. Once you do that, respondents send in the group keyword before any questions are asked; they get a response that says they are enrolled in that 'event'. Then, the codes for individual answer choices within the group are numbered 1-9 and then alphabetically. That is, say the first question in the group has five answer choices; they would be numbered 1, 2, 3, 4, and 5 so students who wanted to select the fifth answer would only have to text in the number '5'. If the next question in the group also has five answer choices, they would be numbered 6, 7, 8, 9 and A so students who wanted to select the fifth answer there would text in the letter 'A'. This can get a little bit confusing since most of us are used to talking about answer choices as A though E but I got used to it. For me, grouping questions together made it easier to keep my poll questions organized. I created one group for each class session, for each section, and named the group with the date and section time; for example, May11AM for my 11am class and May11PM for my 2pm class. As soon as students got to class, they knew they should text in the day's keyword so they would be ready to go when the first question came up. Here's what the first slide of every class looked like (this was up as students walked in):
And here's what a typical question looked like (note that the strip along the side was my 'timer'):
No integration with Blackboard: In order to give students credit for their PE responses, they first have to create accounts in PE and if they want, they can register their cell phones (so any response sent from that phone number is automatically connected to their account). If they don't register their phones, they have to log in each time and submit responses using a browser. Another drawback of PE, relative to clickers, is that it is not integrated with Blackboard, the course management system. This means there are extra steps for students (registering on a separate site) and extra steps for me. To get their daily points into Blackboard, I had to create a 'report' in PE, download that to Excel, make any necessary adjustments in Excel (such as giving credit for right answers versus just participation, or just summing up the points for the day), then upload to Blackboard. A colleague in the business school who also piloted PE this spring has apparently developed an Excel macro that can take care of some of the Excel manipulations but I just did things manually. For me, the extra work wasn't a huge issue but one thing that was frustrating was that in order to upload to Blackboard easily, I asked students to change their identifier to their University ID number (the default when they create their accounts is their email address). By the third week (and after multiple reminders), almost all students had done this but I had two students (one in each section) that never made the change; since I stopped making the adjustment for them after Week 3, this meant that their PE points were zero for every single class and they STILL didn't figure it out! [Note: if PE were used a lot more across campus so this happened in all their classes, I have to assume they would eventually fix it but I'm still amazed...]
Dealing with multiple responses: Another issue I had to consider was how to handle multiple responses. With most clicker system, students can change their responses as long as the question is open and the system will simply retain the last answer submitted. With PE, you can set an option to only allow up to X responses per person or unlimited responses; if you choose to allow multiple responses, PE records every response separately (every response is time-stamped). PE also can send a confirmation text so students can verify their response was received (this is an option you can turn on or off). In my case, since it usually didn't matter which specific answer a student selected, I set things up so they could only submit one response; on the few occasions where their specific choice 'mattered', I made sure to tell students that they needed to be extra careful before sending in their responses since they would only get one shot at it. My colleague in the business school allowed multiple responses and then used his Excel macro to only count the last submission for each student.
Next time, I'll share some of the feedback I got from students...
p.s. While I was working on writing this post, InsideHigherEd had an interesting article on standardization of clickers that mentions cell phones replacing clickers. And if you're more old-school, ProfHacker just posted an article about low-tech alternatives to clickers.
Related posts:
Texting in response to open-ended questions
Student response to PollEverywhere
Low-stakes assessments: I mostly used PE to have students submit their individual responses to multiple-choice application problems before they discussed those same problems in their teams. My main concern was making sure that students had to think about the problem individually a little and commit to an answer before discussion. I didn't care so much what specific answer they chose so students received credit just for answering anything (i.e., participating); there were only a couple times when I made credit dependent on selecting the 'right' answer. PE allows you to embed polls in PowerPoint and that is what I did, rather than switching over to the website each time. One downside of PE, relative to clickers, is that there is no timer so I created one using animation in PowerPoint. It's a clunky workaround but if you want to give students a visual indication of how much time they have to answer a question, I'm not sure what the alternatives are.
Grouping questions: One thing I had to decide was if I wanted to 'group' my questions together or not. The way PE normally works, each answer choice has a randomly-generated unique keyword; for example, if you want to choose answer A, then you send in '70101' and if you want to choose answer B, then you send in '70103', etc. With a paid account, you can also create your own keywords to replace the random numbers (but they still have to be unique since the keyword identifies both the question and the specific answer choice). An alternative is grouping multiple questions together and assigning a keyword to the group. Once you do that, respondents send in the group keyword before any questions are asked; they get a response that says they are enrolled in that 'event'. Then, the codes for individual answer choices within the group are numbered 1-9 and then alphabetically. That is, say the first question in the group has five answer choices; they would be numbered 1, 2, 3, 4, and 5 so students who wanted to select the fifth answer would only have to text in the number '5'. If the next question in the group also has five answer choices, they would be numbered 6, 7, 8, 9 and A so students who wanted to select the fifth answer there would text in the letter 'A'. This can get a little bit confusing since most of us are used to talking about answer choices as A though E but I got used to it. For me, grouping questions together made it easier to keep my poll questions organized. I created one group for each class session, for each section, and named the group with the date and section time; for example, May11AM for my 11am class and May11PM for my 2pm class. As soon as students got to class, they knew they should text in the day's keyword so they would be ready to go when the first question came up. Here's what the first slide of every class looked like (this was up as students walked in):
And here's what a typical question looked like (note that the strip along the side was my 'timer'):
No integration with Blackboard: In order to give students credit for their PE responses, they first have to create accounts in PE and if they want, they can register their cell phones (so any response sent from that phone number is automatically connected to their account). If they don't register their phones, they have to log in each time and submit responses using a browser. Another drawback of PE, relative to clickers, is that it is not integrated with Blackboard, the course management system. This means there are extra steps for students (registering on a separate site) and extra steps for me. To get their daily points into Blackboard, I had to create a 'report' in PE, download that to Excel, make any necessary adjustments in Excel (such as giving credit for right answers versus just participation, or just summing up the points for the day), then upload to Blackboard. A colleague in the business school who also piloted PE this spring has apparently developed an Excel macro that can take care of some of the Excel manipulations but I just did things manually. For me, the extra work wasn't a huge issue but one thing that was frustrating was that in order to upload to Blackboard easily, I asked students to change their identifier to their University ID number (the default when they create their accounts is their email address). By the third week (and after multiple reminders), almost all students had done this but I had two students (one in each section) that never made the change; since I stopped making the adjustment for them after Week 3, this meant that their PE points were zero for every single class and they STILL didn't figure it out! [Note: if PE were used a lot more across campus so this happened in all their classes, I have to assume they would eventually fix it but I'm still amazed...]
Dealing with multiple responses: Another issue I had to consider was how to handle multiple responses. With most clicker system, students can change their responses as long as the question is open and the system will simply retain the last answer submitted. With PE, you can set an option to only allow up to X responses per person or unlimited responses; if you choose to allow multiple responses, PE records every response separately (every response is time-stamped). PE also can send a confirmation text so students can verify their response was received (this is an option you can turn on or off). In my case, since it usually didn't matter which specific answer a student selected, I set things up so they could only submit one response; on the few occasions where their specific choice 'mattered', I made sure to tell students that they needed to be extra careful before sending in their responses since they would only get one shot at it. My colleague in the business school allowed multiple responses and then used his Excel macro to only count the last submission for each student.
Next time, I'll share some of the feedback I got from students...
p.s. While I was working on writing this post, InsideHigherEd had an interesting article on standardization of clickers that mentions cell phones replacing clickers. And if you're more old-school, ProfHacker just posted an article about low-tech alternatives to clickers.
Related posts:
Texting in response to open-ended questions
Student response to PollEverywhere
Hi Jenn -
ReplyDeleteThis is Jeff, one of the founders of Poll Everywhere. Someone on our team told me about your writeup.
I think your analysis of our product is one of the most spot-on, perceptive and useful I've seen (especially as it relates to current product deficiencies). I'm so glad you've found us useful.
In the next year, we'll be working on some things relative to what you've mentioned. In rough order of priority:
- Attendance reports
- Have any keyword; A-E for every poll if you like
- Blackboard integration
- Changing answers
- Timers
I want Poll Everywhere to be much more accessible than hardware clicker companies. Since you've proved your insights to be so perceptive, if you ever have more, feel free to also email us: support@polleverywhere.com or jeff@polleverywhere.com
Thanks for the writeup!
Jeff
Thanks for the info, Jeff! Good to know that these improvements are on the way. Unfortunately, as I'll talk about in my next couple posts, my only real problem with PollEverywhere is something that neither you nor the campus really has much control over - cell phone reception! But it definitely is a great tool that I'll be keeping in my toolbox!
ReplyDeleteInteresting. We've considered shipping a set of 5-10 cheap devices (prepaid phones we know work, or iPod touches / other lightweight web browser tablets) to customers to serve as fill-in devices. It's a little annoying to have to keep track of who has which loaners checked out, but do you think this would have helped those pre-paid and Nextel students?
ReplyDeleteI think loaners would be OK for relatively small classes but my experience with teaching a class of 500 is that if a problem affects 2 or 3 students in a class of 50, I can expect it to affect at least 15 or 20 in the 500-seater, and that seems like too many to have loaners. But if I were going to go that route, browsers would be better than phones, at least on my campus - wireless access is pretty good all over and if students are using a browser, they can just log in and won't have to worry about registering a specific cell number.
ReplyDeleteI used PollEverywhere in my Econ classes this past semester as well. Having a few loaners (especially ones with browsers) would have really helped. I think it was mostly Sprint users who had trouble in my courses.
ReplyDeleteThough most of the experience was good, I did have a few frustrations. One is that the PowerPoint slides were slow to load the PollEverywhere questions even though I had a wired internet connection. I was using the PowerPoint 2003 version, so I'm not sure if the issue would be resolved with the PowerPoint 2007/2010 version.
I agree that adding timers, any keyword (A-E), and BlackBoard integration (or other CMS integration) would be very helpful.
The way I used PE is to have students record their answers to a set of in-class exercises that we would do once or twice in a class period. I haven't ever used clickers before, so I cannot compare between the two. Originally, I thought that if I had the PowerPoint presentation automatically advance the slide after 60 seconds and then to continually loop through the questions, the students would have enough time to answer. But, the students complained that they didn't have enough time to read the question and pick an answer. If I continued with that set-up, it was going to eat way too much class time. So, I resorted to printing the PE slides as a handout, and distributing a set to each group so that they could submit their answers. Then, we would look at the class distribution of answers on the overhead as we went over the answers. This worked much better, though I didn't like having to print so much paper. Also, it would help if I could modify the font and font size of the answer choices more so than is allowed now. Sometimes, the slide handouts were hard to read.
I also think the registration process was confusing for students. Some of them signed up for a monthly plan (as if they were hosting the questions) instead of signing up as a student in my class. Even though I gave them a handout with the direct link on it, some just went to polleverywhere.com to register. Maybe it could be made more clear on the website whether you are registering as a participant or as an organizer. I could also be more clear in my instructions the next time around.
Prof. Savage: We agree that the student registration process is very confusing. We're going to be simplifying this so that the registration link looks something like pollev.com/prof_savage/register. Also, we want the students to be able to login with their Facebook and/or Twitter accounts so that we can go ahead and fill their name and email in for them. We also have a few ideas were students could just register over their phone via SMS so they don't even have to visit a web registration link.
ReplyDeleteInformative piece of writing! Among all the materials presented, audience response system has greater impact in creating positive and interactive learning and interaction outputs. Audience response system benefits broad range of industries and organizations like education, market research, executive decision making, events, conferences, game shows and many others. The use of ARS is not limited to simple PowerPoint presentation. ARS empowers presentation creating effective and interactive setting between audiences and presenters. Audience response system is the latest offering in advance learning, meetings and discussions. It takes you away from hassle-free manual tallying and possible errors. For more info you can visit us @ www.powercomars.com
ReplyDelete