[Evenstar]

Originally Posted by Woodsy

It is truly amazing to me how you just do not get the math... I have quite the grasp on your analysis! In fact I understand it completely and quite possibly better than you do. You used a relatively standard formula to extrapolate data! Not really a big deal, but the formula is flawed in this application and that is shown by the extraordinarily LARGE number of boats. To be clear, when I say boat I meant singular instance as I assumed did you. In no way did I mean to infer that there were somehow 770,400 boats floating on Lake Winnipesaukee!

I “get the math” just fine – and I seriously doubt that you understand this better than I do.

Have you recently (or ever) taken a college level course in Research Methodology? Have you ever done statistical analysis on the data from a municipal survey (that you helped design, and spent time collecting the data on), to aid a town in putting together their comprehensive plan? So don’t be so quick to dismiss me as someone who does not know what she is doing. The problem is that you are still looking at the data wrong – and you are still not getting my conclusions.

And you did post:

” Do you seriously think that there were 770,400 BOATS on Lake Winnipesaukee during those 11 weeks??

Which is inferring “that there were somehow 770,400 boats floating on Lake Winnipesaukee.”

Essentially you have less than 1/3 of 1% chance of coming into contact with a boat traveling over 50! What part of this math dont you understand? I am using YOUR formula! No matter how you extrapolate & multiply the numbers, the ratio/percentage doesnt change.

My chance of encountering a high-speed boat is way higher than what the static snapshot and you calculations suggest – because all the boats are moving – and what I didn’t factor in (and could have) were the boat speeds.

For instance, a boat traveling at 50 mph uses (or covers) 10 times more of the lake than I do in my kayak at 5 mph – during the same time period that we are both on the lake. So the faster that a boat is traveling, the more likely it is that I will encounter that boat on the lake – the percentages increase directly as the speed increases. The lake is roughly 20 miles long, so if I paddle from one end to the other it will take me about 4 hours. In that 4 hours a boat that is covering the same 20 mile stretch of lake at 50 mph will pass me 10 times.

This represents 10 high-speed encounters in just a 4-hour period of time, with just 2 boats on the lake (who happen to be using the same 20 mile stretch of the lake). Those 11 weeks equal 770 daylight hours. Therefore my chances of encounter a high-speed boat are considerable – which is what my actual experience paddling on the lake has been.

Now... on to the flaws with your formula! By using YOUR formula we can extrapolate that the NHMP would have clocked 770,400 boats over the 11 week survey. Good luck trying to get ANYONE to believe that number!!! There is no way that number could be remotely accurate and here is why… The reason your formula/logic is flawed is because you are taking a blanket approach. If you knew the lake better, perhaps you would understand the flaws in your formula.

As far as your points 1 through 4: I clearly stated that my analysis only covers the 11 weeks of the study – that’s what the 770 hours represent.
I also clearly stated: “If the study was done correctly and the “snapshot” was an accurate representation of the speeds of boats on the lake, than my statistical analysis is just as accurate.” My conclusion is and has always been that this study was not done correctly – and that data which was collected is flawed. And once anyone takes the data and tries to do statistical analysis on it they get some really strange results.

The problem is that the data collected does not accurately represent the average boat traffic on the entire lake over the entire 11 week period. The MP purposely used areas of the lake where boat traffic was high in order to record a large number of readings over a relatively short period of time. In 135 hours, they RECORDED the speeds of 3852 power boats – that’s more than 28 boats per hour or 1 every 2 minutes! And this is just the boats that they were able to record, not the total number of boats that went by them.

My contention is that, by using the busiest areas of the lake for the study, the study did not represent the average condition on the entire lake. Boats are going to travel at slower speeds in areas where the boat traffic is high.

5. All boat traffic on the lake is not equal... so your blanket approach of multiplying to cover the entire area of the lake is inherently flawed. For example areas such as the Weirs and Meredith see alot of boat traffic, while areas off the beaten path such as Moultonborough Bay and all of the little coves see very little boat traffic.

There is nothing wrong with my approach – that is exactly how statistical analysis is done. But (and this is a really important but) any statistical analysis is only as good as the data that was collected.

The speed study is being used as “proof” that few boats on the lake are traveling at speeds over 45 mph – so there’s no reason to enact a speed limit. What it actually shows is that on areas of the lake where boat traffic is the highest, there are only a small percentage of boats that are traveling over 50 mph on the busiest times of the highest boat traffic days.

Had the study areas actually represented the average conditions on the lake, on an average day during these 11 weeks, the data collected would have been totally different. And statistical analysis of that data would have given very different results. But then the data may have supported the need for a speed limit - which is why the study was not done correctly.

6. There is no reason to adjust the raw data at all.

Yes there is. Even the report is called “RECREATIONAL BOAT SPEED SAMPLING.” A sampling results in raw data that is supposed to accurately represent the overall population/area/time period/condition. Since boats were not recorded over the entire lake, but only in sample areas, the raw data needs to be adjusted to cover the entire lake. And since the data was collected only during less than 2% of the total daylight hours of these 11 weeks last summer – the total hours need to be factored in as well. This is what statistical analysis does.

I see the forum sharks are still circling. I'm sooo impressed . . . it takes . . . let's see . . . 5 of you guys (at this point) to debate (or to try to intimidate) just one college girl.