Thursday, May 27, 2010

(Possible) SIP Proposals

One possible SIP topic using publicly-available data could be analyzing the evolution of the health care debates that have transpired (which is not to say that such debates are even close to being fully finished). While this SIP would obviously deal more than with simple statistics, such data would prove essential to any argument that would be made. In order to study such findings, one would compile a large number of surveys measuring public opinion of health care or other such factors of certain aspects of health care reform (i.e., the approval of a public option). Then one could analyze these results over time and measure whether there were statistically significant findings with regard to approval over time. In addition, further hypothesis tests could be done to see which demographic groups most strongly supported/opposed health care reform and its various aspects (such as the rich/poor divide or gaps between races).
Furthermore, additional data could be looked into such as if there is any correlation between support for health care reform and general trust in the ability of Congress to pass legislation. Data could also be used to analyze the voting patterns of congressional members by viewing the proportions who voted in line with their political party or whether an impending election (2010) may have had any influence on voting patterns.
Such analysis may then possibly lead to a conclusion as to whether Democrats could have simply strong-armed a health reform bill through congress that included a public option rather than the admittedly watered down “negotiation” bill that has passed. Of course, numbers cannot fully tell the whole story but it would offer great insight to the power of the minority that is now found in the legislature.

These data are easily found on the internet through various means:
Polling Sites:
News Sites:

Another SIP topic would be one where one would collect original data in order to measure the effectiveness of Woodward tutoring programs (including both the after-school PALS program and regular in-class tutors). By partnering with the school, one may be able to obtain data from standardized testing and be able to compare the scores of students who have had no contact with tutors, contact with either in-class tutors or a PALS tutor, or a mixture of both. One would then be able to test a hypothesis test in which one could measure the level of significance to which the tutoring programs have assisted (presumably) in raising students’ scores. With large enough findings, it may allow the programs to receive more money so that they are better equipped to help assist such an underprivileged school demographic. Furthermore, tests could measure the differences seen between boys and girls, income, etc. in order to analyze further socioeconomic factors that a simple test of tutor/no-tutor would leave unexplained.
Ideally, the sample size would be able to be the whole Woodward Elementary student population since obtaining all the data as opposed to just some of it should be a negligible increase in difficulty.

No comments:

Post a Comment