![Workshop Goals ,[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]](data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7)
More Related Content
Similar to Jg Ak Vid Wkshp Vsa Pt1
Similar to Jg Ak Vid Wkshp Vsa Pt1 (20)
Recently uploaded
Recently uploaded (20)
Jg Ak Vid Wkshp Vsa Pt1
- 1. Assessing Learning in Process: Videotaping Visitors in Museums Josh Gutwill Acting Director of Visitor Research Adam Klinger Evaluation Systems Specialist
- 6. Adam Klinger GIVE: a Lab Study
- 8. Adam Klinger GIVE: a Lab Study
- 9. Adam Klinger GIVE: a Lab Study
- 10. Adam Klinger Geometry Playground: Eye in the Sky
- 17. Adam Klinger APE: Recording at single exhibits on the floor
- 20. Adam Klinger APE: Recording at single exhibits on the floor
- 23. Adam Klinger APE: Recording at single exhibits on the floor
- 25. ย
Editor's Notes
- Welcome to this workshop, called Assessing Learning in Process: Videotaping Visitors in Museums Iโm Josh, this is Adam We use video to study visitors
- Make this quick, then go over agenda
- A better vantage point for observations You can put the camera high overhead, or even directly at peopleโs faces in a way a live observer could never do The ability to hear visitor conversations A live observer would have to standing almost on top of them to be able to hear as well as a well placed microphone Even then, it is difficult for someone to get a conversation word for word A way to reliably capture subtle phenomena You can play video back again and again to catch what you might have missed the first time Better interrater reliability Coders all see the same video and hear the same audio- getting agreement is always difficult. Even when there is disagreement, coders can review together and resolve it Visual evidence to share with stakeholders Thereโs nothing better than seeing and hearing yourself to be convinced of something or understand itโs impact
- When do other methods work better Setting up cameras, mics and cables is time consuming, and not appropriate for quick studies with small Ns. Coding video takes more time and resources than coding interview data, so itโs not appropriate for timely feedback on prototypes Capturing, managing and coding video data requires technical expertise, so itโs not appropriate if you donโt have resources to support that kind of work
- Video systems use different components and configurations depending on the study goal The Exploratorium has used many systems for different projects AndIโm going to show you examples of some of the systems weโve set up Each system has its advantages and disadvantages Maybe one day โweโll have something put togther that is flexible enough to do everything we need, but for now weโve had to set up separate systems Focus on video recording methods In these examples Iโm going to be talking about the recording technology, Josh will discuss informed consent issues later Iโm going to start with a system for a very controlled environment and move on to systems for more real world
- GIVE Lab (queue photo of lab space) The GIVE study was a rigorous randomized controlled trial where we developed different pedagogies for teaching inquiry skills and compared their effectiveness in a pre-post design. Because we wanted to control as many aspects as possible, we built a specialized lab, off the museum floor. This made good audio and video recording a relatively easy proposition. A controlled environment allowed us to: reduce outside noise position exhibits and cameras in such a way that we had a clear viewpoint. control lighting through trial and error, we were mostly able to optimize audio quality before we collected actual data. . We sent signals from four cameas [view of cameras if we have it] to a recording station in another room. As visitors were only at one exhibit at a time, we simply used a simple switch to move from one camera to the next. Although there were some problems with exhibit noise, all in all we were able to get excellent audio and video with simpler equipment and little fuss
- A randomized clinical trial design The GIVE study, stands for Group Inquiry with Visitors at Exhibits had two control groups, two treatment groups and a pre-post assessment A lab was built and used for the study in order to control for as many extraneous variables as possible Having a quiet, well lit place to work meant that getting good video and sound was easy Cameras and mics were permanently mounted above the four exhibits Quiet conditions in a small space made getting good quality audio relatively easy Even in this environment, we had some troubles with exhibit noise, but overall, quality was good, even with Omni directional microphones
- Hereโs that shot of the room again, and you can see The mics hanging over the exhibit, Cameras are mounted high on the wall, and all the signals are running to a computer in the interview room. Since visitor groups only used one exhibit at a time, it was easy to switch signal from one camera to the next in a single recording
- So now I want to show you a couple of minutes of video from the study here. This was recorded with only two mics, neither of which was especially close to the visitorโs faces You can still hear the roar of the crowd coming through the wall, but all and all the sound is good
- Next I want to show you a system that we used with a current project, Geometry Playground that places exhibits in a large open area on the museum floor- this a shot take the other day of the current set of exhibit. You can see thereโs a lot going on there
- An exhibit development project with a research component Studies families as they use multiple exhibits Parent / child dyads are recruited and recorded as they move through the area
- So we recruited visitors and asked on kid and one adult to wear these clip on mics, and a a mic recevier thatโs about the size of a deck of cards that you can see on the table here. That signal is picked up by a nearby receiver like this and sent about 300 feet back to the evaluation office
- The video is recorded by two Pan Zoom Tilt cameras mounted about 35 feet over the area, theyโre almost unnoticable unless you know to look for them These cameras can pull back far enough to see the whole open area, or zoom in to see a small item a visitorโs fingers. In
- GP eye in the sky So both sound an video go back to a lab where a researcher uses joystick like those on the right to control the cameras and the two mics and 2 pictures are synchronized and recorded using security system software. It takes a bit of practice to track two visitors at the same time, but itโs kind of fun, very video-game like
- This system works really well for itโs intended purpose Wireless microphones allow us to: Hear visitors clearly even in noisy areas PZT cameras allow us to: Track visitors through the area, even if they separate Choose the best angle to see whatโs important Use one camera for details and the other for the big picture
- Now I want to show you a short sample of a parent/child pair being observed using two exhibits Cue circles. Weโre recording a man and his daughter, that are in the red circles. They also have a smaller boy with them, who was not wearing a mic but can be heard in places
- GP eye in the sky Geometry Playground is an exhibition thatโs currently in development at the Exploratorium. The project has a research component in which we want to be able to analyze parent/child dyads as they move around the exhibition. This requires that we be able to see and hear visitors wherever they go, even if they separate from each other To get good sound, we are using wireless microphones that are pinned to their shirts. Visitors wear a pack with a small transmitted that sends audio to a receiver mounted nearby. [switch to photos of above cams] To get clear video, we mounted two PZT (Pan/Zoom/Tilt) cameras about 35โ above the museum floor on either side of the exhibition. That allows us to: pick the best angle to watch them from at a given exhibit track both adult and child if they separate use one camera to focus on what theyโre doing with their hands (e.g., rotating a shape to make it fit together with others) and the other to give the coder a sense of the overall area- crowdedness, other activity, etc. [show tracking vid] [show Karen driving]The cameras were controlled and recorded by an Exploratorium researcher from a computer in the VRE office, about 300โ away. The fact that people that had not signed consent forms could inadvertently be picked up by audio or video required a complicated setup for ensuring informed consent. Again, Josh can talk about it later Bottom line- audibility is excellent, and with much practice researchers were able to get high quality video that was able to pick up nuances, like manipulating a geometrical shape in the hand, while also preserving big picture info, like crowdedness and path through the exhibition. ย On the other hand, asking visitors to wear a fanny pack is a lot to ask, and combined with signing consent forms is likely to make them hyper aware that theyโre being recorded
- Interested in real-world experiences We wanted the area to look to visitors like any other part of the floor We wanted visitors to use exhibits uncued Needed video of many visitor interactions at a single exhibit We wanted to capture conversations between visitors
- Phase 1: recording in an open area of the floor We used sturdy telescoping stands We used two shotgun microphones Iโll talk more about shotgun mics later and youโll have a chance to try using them, but basically they pick up sound where theyโre pointed and not other places. This means that they potentially can get more of the exhibit userโs voice and less of the background, but also require careful placement
- Now hereโs a quick sample of audio recorded this way This exhibit is called downhill race, which is actually about 8 feet wide which is a lot to pick up with two shotgun mics CUE MIC CIRCLES So we placed them at the top and bottom of the ramp where visitors spend most of their time using the exhibit. Their conversation is relatively easy to understand, but you can hear thereโs still a lot of noise in the background.
- Phase 1I: Building the Sound Abatement Area Although that setup you just saw was much better than things tried in the past, we still needed to get more consistent sound, which meant creating a dedicated Sound Abatement area. Still part of the floor, this area has several changes to help the audio Sound absorbing features Cutting the ambient noise level down was the first priority. Carpeting The carpeting extends beyond the edge of the area to keep sound from bouncing off the concrete floor into the space Sound absorbing walls Absorb sound, cutting echos Sound absorbing hood Same 4 shotgun microphones We also added another pair of shotgun mics, now 4 in total, mixed together with a ceiling mounted mixer
- Quick setup features: While we were at it, we also added some features to make prep for recording easier. Trusses on ceiling Movable clamps to mount: Microphones Camera Mixer With these in place we can put the cameras and mics almost anywhere weโd want them and do it quickly.
- Hereโs the last video sample taken recently in the Sound Abatment area
- If youโre interested in the way these systems were wired together, there are three schematic diagrams in your packets- one for the Eye in the sky, and one of each of the APE project systems for recording individual exihibits Later today weโll be using these as a guide for setting up our own systems .