User:Dante Anton
ANTH174_10 Dante Anton
Anthropology Major, Senior
I wouldn't call myself fluent with computers or coding, but I'm looking for a challenge, which is why I'm taking this class.
Contents |
Links
Day One Notes
Thought provoker: Historiography is an interesting concept. Although we as a group live in the present, essentially living in and writing our own history, which will eventually be seen as the past or "factual", the recent advent of global networking and information technologies and our subsequent dependence on those technologies gives the "right" of history to anyone with an internet connection and a drive to revise it as he sees fit (see Wikipedia). Of course with such editing comes consequences; however; enough money and power is all a company or government needs to cover their tracks. Doug 11:50, 6 January 2010 (PST): Yes, but see Surreptitious wiki edits, which exposed 1000s of companies up to such tricks, as published and exposed in the NYTimes.
To create a link simply surround the hyperlink with double brackets. Copy the variable codes in a previously made program and insert your own variable to start your project
hashmarks # or ## or ### are used as a way to signal the program to not code the line
double equal sign (==) surrounding Text creates a header.
Ctrl+A copies all text in page
Day Three Notes
when you edit a page, all you will see is that specific page. Add your name to the EduMod page with the number that you want to prevent others from taking over that specific site.
- I am number 66
P value-what is the probability that the the results are simply random Unrestricted, or XUR, differs from restricted as it incorporates all variables negatively correlated results, such as female power in relation to polygamy, means that the prevalence of polygamy is negatively effected by the variable of female power use a star to add a space at the beginning of the page
Day Four Notes
Reminder: sign up for Friday's conference at the third floor of Anteater Instructional Bldg at room 3030 @ 1:30pm-3:00pm...getting easy extra units (and enjoying the process) is always a good idea! Class Code 71500
Follow the Checklist for the Class to make sure you are progressing with the other students.
- new Items on the Checklist: Choose a new dew dependent variable, run the variable through the program check results and see if any correlations arise.
Limits of Causality- When finding correlations and Equivalencies between an independent (X) and dependent variables (Y), one may come across an implausable causal relation in which X is unlikely to cause Y, but Y may cause X.
- ex: aggression in teenagers causing a higher occurrence of warfare in a society (or Presidents less endowed with civility and reason) and visa versa vs. Mud does NOT cause rain, but rain MAY cause mud.
Avoid These limits of Causality by carefully choosing dependent variables to focus on while simultaneously keeping the number of dependent variables to a manageable level...something close to the dependent variables used to run our first program on R.
Professor Doug White's Presentation Notes:
- SCCS data base; Judea Pearl had a very strong influence on the project
- Current class aims at further designing The eLearning Open Repository for Anthropological Knowledge (eLorak) with the intention of one day having it open to all classrooms with internet access
- Benefits of SCCS:
1. Take into account notions of possible causality.
2. Benefits to Data and accessibility, both civil and research based.
3. Findings in the models will form multiple networks.
Solving Galton's problem of nonindependence using 2-stage least squares to improve regression analysis that set the stage for Bayesian causal analysis.
- detecting causality i.e. low p-value from the acquired data
- imputing missing data; Donald Ruben predicts missing values for different configurations thorough taking best described dependent variables settings like climate(called auxiliary data base)and uses probability to create these values. Theses predictions build unique data sets, which in turn can be combined to create one complete set.
- instrumental variables are used to correct concentrated, patterned errors, which should not occur in a data set. Essentially takes observations to create an experiment and plays a roll in identifying and estimating causal effects.
- 2-stage least squares (2SLS0: instrumental variables in 2SLS reduce the significance by clustering the effective sample size into groups. Weighted matrices and matrix multiplication are used as the physical process to find the prediction.
Detecting Causality with Bayesian Causal graphs
- independent var X has a weighted correlation to depvar Y which can be graphed.
- Multiple imputed (MI) missing data are taken into account by instrumental variables and auxiliary data base.
Day 7 Notes
make sure to get ride of factors using delete arrows, versus keep arrows, with a P value of three or higher.
pound signs exclude lines from coding. ##comment out everywhere yoru dependent variable appears on the project
define independent variables by SCCS$vxxx
make the unrestricted to restricted model, run it, copy the results. copy and insert your own variables figure out what plus signs do.
This is what I'm going to use
1721. NUMBER OF RICH PEOPLE (WEALTHY)
88 . = missing data
27 1 = absence of rich (original code 10)
41 2 = presence of rich, no information on numbers
(original code 20)
27 3 = few rich (original code 21)
3 4 = many rich (original code 22)
never change xUR ecept when there is an indepver that completes with the depvar. add indepvar after your xR is DONE
Notes to self for powerpoint
- Use courier font to make sure all tables are aligned.
- Pay Close attention to +/- values.
- Include copy of last restricted prog results.
- Clean up EduMod page.
- 1723 Poor is positively correlated with Wealthy. Find results for both xR and xUR to compare.
Notes to self for final paper
need to take into consideration negatively correlated indepvars "gath"SCCSv203 and "pathstress" SCCSv1260
