Blast it. This is like an itch that won’t go away.
OK. If I.m getting this, you’re going to perform some sort of trend analysis on the state legislatures and their members. All 7000 as data points? Presumably looking for a statistically significant trend, at some sort of confidence interval.
A couple of questions: methodology for sampling? Absent a vote, I would think it difficult to pin down a number one way or the other and a vote is usually the end result of a process. You have to have a bill or resolution introduced, which for whatever reason might not happen (deadlines being a particular constraint that comes to mind, or leadership priorities)
Methodology for analysis? I’m not a statistician so I don’t understand in depth the plusses and minuses of the various distributions. I assume a standard normal distribution?
Thanks again (i think) for a provocative and insightful post.
I have a database of all 7,383 state Legislators, their phone numbers, emails, contact info for assistants and some of the legislative geographical boundaries (working to complete), party affiliation and political views (conservative, liberal, moderate, uncategorized).
I’m also adding to the database which year the state legislators were elected, how many terms they have served and whether they ran or are running unopposed (as many as 30 percent!), and their salary levels as those levels are quite variable and affect decisions to travel to COS Project events.
I can email the more than 10,000 legislators and assistants in the database and poll them on their willingness and plans to attend COS Project meetings if they haven’t already, if they favor states rights issues using Article V, if travel expenses are an impediment to their attendance in COS Project meetings and if there is a PAC for their travel subsidies and what is the contact info for such a PAC if it exists.
Because I email all state legislators and their known assistants, I have no sample design. However, I do use statistical methods to adjust for non-responses which are important to non-biasing in survey sampling. Non-responses are emailed weekly over a period of one month before they are declared as actual non-responses. Response rates are excellent at the state legislator level.
I have a short list of 5 questions that are asked on the email survey and the questions do not change for those legislators that are uncertain. I can measure their change using a simple Likert scale over quarterly surveys. I also leave a professionally recorded voice attachment to the emails as well as a link to Mark Levin’s superlative speech given last month before ALEC at the COS Project meeting in Mt. Vernon. The surveys are standard attachments and are aggregated using a series of spreadsheet macros.
I am working on an amendment that I am hoping will appeal as natural and appropriate to State Legislators and incorporates a good portion of Mark Levin’s Liberty Amendments.
Here’s a link to a proposed Amendment 28. Tell me if you think it would appeal to state legislators. Your feedback is appreciated.
http://www.freerepublic.com/focus/news/3244067/posts?page=38#38
I am also working on a campaign booklet for persons such as Freepers to use if they can be persuaded to run for a state legislative office. It details how many supporting activists they will need, what kind of funds and organization needed to win and tips on assessing the probability of winning in the state districts they live in.