Current Count 79 participants.
Please, if you can PM your friends on the board or contact others with DP/DR symptoms you should send them to the study link. PLEASE DO NOT TAKE THE SURVEY TWICE! This will only do a disservice to our data. However, please encourage as many individuals that you can to take this study. Our first goal is 200, but 500 would be a very strong number and I believe we can reach that goal.
Dr. Simeon is pleased with the response rate so far, actually her last e-mail regarding the response rate simply read: "Fantastic!"
However, please take the time to take the test. I can not give any preliminary results, however interesting data already is beginning to form. However, in order to make data significant, you have to show statistically that the data you have is not just caused by random chance, but instead by a "statistically significant" chance that it was caused from a real pattern/relationship/etc. For example, if we have 10 people toss a coin and 7 of them get the Front side, there are mathematical calculations to determine how likely it is that this result could happen at random. With only 10 toin tosses, having 7 be heads (which is 70%) would NOT be statistically significant. However, if we had 10,000 people toss coins, and 6,000 of them had them land on the Front side, we would have 60% of coin tosses being Front and 40% being Back. BECAUSE the data set is so large (10,000 tosses), this is now EXTREMELY significant. One would expect a 50/50 ratio, however with so many tosses the "odds" of having 60% of them being the Front side is considered VERY significant.
Essentially, a value is assigned called a "p" value, which is generally expressed in the terms: p = x, where x is the chance that this result would happen at random. So, if you calculate your data and receive a p = .95, then there would be a 95% chance that the hypothesis you were testing could have happened randomly. However, if your p = .002 then there would only be a .2% chance that this result could have happened at random. The P value is strongly affected by the NUMBER of DATA points that you have. Tossing a coin twice and having it turn up on the same side in a row is not signiciant, but toss the coin 100 times and have it turn up on the same side -- now you have some significant.
I do not want to go in to more detail about the statistical analysis of data, but just want to say that the more data the easier it is for us to draw conclusions that we can express as being signicant and not random. In this case, if 90% of 3 individuals said that Gabapentin improved their symptoms we could not publish this data and say that it is significant. However, if we haad 500 participants and 400 said that Gabapentin improved their symptoms, then we could use this to form an argument that we should study Gabapentin (generic for Neurontin) as an treatment for DP/DR. We use this data to get funding for a study on the treatment of DP/Dr with Gabapentin (and also similarly related anti-convulsant drugs).
Because this is an internet survey, we are going to be scruitinzed even more for our data because we do not have a way to validate if you are lying or not. However, the expected "noise" created by individials just taking the test and selecting random answers is reduced by the amount of individuals taking the test. Also, it helps when you write a detailed story on the last question, as this demonstrates to us that your data is real and that the information contained in our database is a real person who is affected by DP/DR.
My apologies for rambling and poor grammar, however I must get to work before I am late.
Best,
David
