What would you ask a researcher about Lake Champlain? If you're anything like the typical visitor to ECHO, you'd ask about something else. (Yes, I have had questions literally asking about "something else".) We give visitors 3 different topics that they can ask researches about, including Blue-green algae (102 selections), Invasive fish (73 selections), Phosphorus or Nitrogen (62 selections). If none of those match their question, they can select a 4th option - "Something else" (520 selections).
Of course, if you've read my previous post and done the math, you'll notice some inconsistencies in numbers. How can we have 757 selections, but only 269 videos recorded? Your response is saved as soon as you select it, but you have many options to abort recording your story afterward. If I filter it down to only people who have recorded a video, we have 25 responses for Blue-green algae (of which we kept 6), 21 responses for Invasive Fish (of which we kept 7), and 18 responses for Phosphorus or Nitrogen, of which we kept 4. The remaining 205 responses all chose "Something else" - and we kept 56 of them.
So, what can we tell from these numbers? Either we aren't offering the right tags, or visitors are not selecting the correct tags for their videos. As this Kiosk is meant to be an important prototype for our larger exhibit, I feel it is important to correct this issue while we still can test. After all, a map where the only way to get any videos to pop up is by using the tag "something else" isn't going to work.
I've used the YouTube channel to gather some additional data on the problem. I chose this, rather than trying to do an exhaustive analysis of all videos recorded, because a) it was a lot easier, and b) well, did you see reason a? The YouTube channel contains two things. Good videos, and Champ videos. So, I've taken the last 20 videos answering the "What would you ask a researcher about" question, and analyzed the response for tag accuracy.
Tag - Topic
Something else - Question about Champ
Something else - Researchers should communicate stewardship to boaters, not just farmers.
Phosphorus or Nitrogen - Fertilizers containing Phosphorus and Nitrogen make our local lake smell!
Something else - How many types of fish are there, and have any gone extinct from Lake Champlain?
Something else - How old is Lake Champlain?
Something else - Are you studying if Champ is real?
Something else - Is Champ real?
Something else - Is Champ real, or related to a Sturgeon?
Something else - Is Champ real? What is he?
Something else - How polluted is the lake in comparison to other lakes?
Something else - How long does Champ live, and how many eggs the female lay?
Something else - Does Champ live at the bottom of the lake? Is he Blue?
Something else - How long has Champ lived in Lake Champlain? What does he eat?
Something else - Why is the lake blackish, and not blue?
Something else - How did the Champlain Islands get their names?
Something else - How much pollution could you clean up in one day?
Something else - Do you ever plan on going on a really big search for Champ?
Something else - We love Vermont. Honor your ancestors.
Something else - Asking about Champ.
Something else - Do you ever find sea creatures in Lake Champlain?
So, what can we determine from this data? Well, as you already knew, everyone wants to talk about Something else. But what are they talking about? Half of them are asking about Champ. Pollution appears to be another theme, and while they don't all show up on YouTube, I can assure you that more than one person has asked how old Lake Champlain is, and how many different species can be found in the lake.
The next step was to filter this data into a time period. Between July 6th and July 27th (the past 3 weeks), 9 people asked researchers about Champ, and 20 people recorded videos about Champ.
So, with all this data in hand, the great experiment can begin.
Every time we get a Champ video, I have to do a few things. First, we remove it from the Kiosk. These aren't the kind of questions we're looking to gather for researchers to answer. Then, I have to remove the videos from the VFL playlist, and manually add it to the Champ Playlist. Why not let the visitors do this themselves? I could change the script to place videos tagged Champ onto the Champ playlist automatically. I could even set the kiosk to hide those videos by default, so I could review them less often (For example, each Monday morning, or Friday afternoon).
Based on this, I have decided to add a Champ tag to the list of possibilities for the researcher question. Over the next three weeks, I will evaluate what happens. Do the tags shape what people are talking about? Will we get more Champ videos than in the previous 3 weeks, because people see the tag and think "I want to talk about Champ!"? How reliable is visitor tagging? Will visitors use the tag? Will visitors misuse the tag?
Check back a little after August 17th to find out!
~Travis Cook, Information Technology Coordinator