I'm taking a break from this year's election to write about my experience as a presenter at PodCamp 7. This year I presented on the same topic "Presenting Statistics in Social Media" with modified slides which can be seen above. I had a larger crowd this year with 14 surveys completed compared to 12 last year. I did not count how many did not complete surveys this year but there were a few. Just about everyone completed surveys last year. Mike Sorg, one of the organizers of the event, told me that "if everyone stays for a session that means you are doing well". While I was glad people stayed for my session, the survey, which could be filled out anonymously, did give me some enlightening comments that I might not have received face to face which I will summarize below.
The slides I presented are represented above. Plus I presented a video that I was unable to show last year. Which you can see here. I also demonstrated how Facebook and Google Analytics presents statistics. I will post the full video of my presentation when it is available.
First I will compare last years responses to the closed ended questions to this years. The questions were exactly the same. The first five questions were coded from 'strongly disagree' to 'strongly agree'. For question one no one said they strongly disagreed that the slides were easy to follow for either year. Though two disagreed with the statement this year, the overall distribution of responses was statistically the same (p=0.667).
Question 2 was a referendum on my clarity as a speaker. Though there seems to be more agreement on my clarity this year there was no statistical difference between last years ratings and this years (p=0.176).
Question 3 relates to how the graphics were presented. I was able to present a video from a BBC documentary called the joy of stats. There were no large overall differences from this year to last (p=0.667).
For question 4, on whether they learned any new information, the responses seemed most identical across years with a few more disagreeing this year (8.3% in 2011 and 21.4% in 2012, p=0.347).
Question 5 is an overall assessment of the presentation. Last year someone strongly disagreed that it was helpful. This year I'm glad no one did but there was still no difference in the overall distribution of responses (p=0.820).
The sixth question relates to how comfortable participants were with statistics. This is the only question where respondents expressed a statistical difference between this year and last (p=0.036) with this year expressing less comfort with the subject.
The last closed ended question was just about whether they had taken a statistics class before. Last year nine respondents had, two hadn't and one didn't answer the question. This year they were split 50-50.
This year I received nine comments on 14 surveys compared to three last year. Those who took stats classes seem just as likely to comment as those who did not. For those who wanted more information on Google Analytics I'll direct you (if you see this post) to Katie Vojtko's presentation on "10 Reports Your Boss Will Love" with the slides and video of it given below.
With the election coming up social scientists will go over exit polls in a similar manner as I went over this small feedback survey. Now that I have a paper trail it does make it easier to double check the numbers. This is the first step in the process that Nate Silver uses to forecast the election today. When the video of my presentation is posted on the PodCamp YouTube channel I will post part 2 when I'll respond to other comments. Tori Mistick in her presentation said that facebook pages get the most views if you post between 1 and 4 pm between Monday and Thursday so that is what I'll try.