1) it is the most recent published report and most relevant to the situation we see today. The most current report should represent the most polished methodology of all the reports as ideally they would learn and adjust from any mistakes/errors in previous reports. It would be silly for someone like me to criticize them for mistakes in older reports that they have already corrected for in more recent reports.
2) Going through the reports takes time and I only have so much. Reviewing several reports would take considerably more time and unfortunately my real job has priority. I have made no remarks concerning the validity of previous reports and the assumption of validity is represented in several of my recent comments regarding trend data. If I am wrong in that assumption, please let me know.
Regarding your last point, that I "screwed up all over the place:" please be specific. Other than a minor math error which had no significant effect on the conclusion of my analysis, what errors have I made? I believe I have provided enough information for an informed, critical dialogue beyond a simple "YOU ARE WRONG!" I took the time to critique the report. I was specific, detailed, and open concerning my analysis. If you are going to critique my analysis, I simply ask that you do the same. I would love to be 100% accurate in every little thing. Unfortunately, I only have so much time to double-check everything and I am only human. If there are as many mistakes in it as you indicate there are, I welcome the knowledge as I absolutely despise disseminating misinformation. Please use more than one sentence in your critique as I would like to understand how you believe each potential mistake specifically impacts the significance of my overall analysis. Basically, please give me more than: You made mistakes, the whole thing is crap!
Last edited by CycloneWanderer; 09-25-13 at 09:00 AM.
If you don't have time to do that, then you shouldn't claim that they are wrong. It is dishonest.
And as far as detailing your mistakes, I have already done so. Do I really need to explain it as:
1) You left out the #'s for 2009
2) You left out the #'s for 2008
3) You left out the #'s for 2007
and so on and so forth back to 1993?
And as far you "taking the time", you just admitted that you did not take the time to do the job correctly
The design of the study is a longitudinal repeated measures survey. You see that last data point? 23.5%? That isn't an aggragate data point from the last 15 years, that point represents the data from the latest survey period (2010). If you read the report thoroughly, it clearly explains that graph. Additionally, from an experimental design standpoint, rolling aggragate data of the type you seem to think this is is virtually useless in this type of longitudinal study as it would actually hinder people from being able to accurately and quickly percieve trends in the data (it would harshly skew the data). The data represented by that 23.5% data point does not include any survey results from before 2010, so I do not need to look at the specific data from other years to be analyze the scientific validity/reliability of that specific data point. In this case, the data point is neither reliable nor valid as it is impossible for anyone to accurately compute the same number given the specific data provided in the report. Honestly, I was hoping your critique would be a little more informed (a little less ridiculous?) considering the level of vitriol you have displayed.
I find it ironic, considering your harsh yet utterly baseless critique, that the following sentence is in your signature:
"One can only be so intelligent, but stupidity knows no limits"
Last edited by CycloneWanderer; 09-25-13 at 01:08 PM.
Data collected at different points in time can only be validly statistically aggragated when there are no significant differences in the data collected caused by the difference in the time of collection. Such a procedure is typically antithetical to longitudinal research unless you are looking at a difference cause by a distinct event at a distinct point in time. Even then, you would only aggragate the data into groups before and after the event prior to comparing those two groups of aggragated data. You would never aggragate the entire data set.
Last edited by CycloneWanderer; 09-25-13 at 01:23 PM.