Anthro 174AW : Thurs WEEK 4


AS BEFORE, dont forget that by Tues: you need to "Turn in one additional page with one or two tables supporting or not supporting one of your hypotheses, and taking into account POINTS 2 AND 3" from Tuesdays' assignment sheet.

FOR TODAY - Regional Replication is pretty easy using SPSS (or pre-windows MAPTAB).

Using region (variable 200, recoded for Old World/New World) as a control, try to replicate the cross-tabs for a pair of variables that are relevant to your hypothesis, whether or not they correlate strongly. Alternative: try using Korotayev's new variable on world religions as a control. These apply only if you want to merge the religion variables into the main datafile

  • F)ile O)pen your main datafile *.sav in M:/Anth174 --
  • D)ata M)erge V)ariables -- and pick the Religion.sav dataset.

    Here are the main instructions:

    1. In Spss:

    • T)ransform R)ecode I)nto different varibles, chose region (v200) as your input variables, name your Output Variable OldNew, click O)ld and New Values. Then enter in the Old Value box and the New Value box the pairs of old and new values as below, and after each pair is entered click A)dd so that the following pairs entered show in the recode window 'Old --> New' variables (not to be confused with old/new world):
      2 --> 1 (thus Circum-Mediterranean is recoded '1' for OLD WORLD)
      3 --> 1
      4 --> 1 (you might want 2 not 1 here if you want to treat Insular Pacific as a separate region '2')
      5 --> 3
      6 --> 3 (thus North and South AmerIndians are recoded '3' for NEW WORLD)
      Finally, click C)ontinue and OK and your new variable will be created
    • S)tatistics C)rosstabulations -- 1 variable for row, 1 for column, and v200 OR BETTER, YOUR NEW RECODED VARIABLE for the third variable at the bottom that will be used as a control.
    • Print your outfile if you want to keep a copy.


      These apply only if var 200 is not in the main datafile and you need to merge
    • F)ile O)pen stds09.sav -- that is the file that has variable 200 (region).
    • D)ata M)erge -- and pick the dataset for one of your chosen variables.
    • D)ata M)erge again -- pick the other dataset for your second chosen variable.

    2. How to interpret your 3-way table with 200 (region) as a control?

    • The resulting six tables will be cross-tabulations of your two variables, one table per continent (1=Africa, 2-Circum Mediterranean, etc.)
    • If one of the six tables doesnt have enough cases or one of the variables doesnt vary, it can be ignored for purposes of replication. (It doesnt count either for OR against your hypothesis).
    • The remaining tables may show a relatively consistent pattern or an inconsistent pattern.
    • Some may support your hypothesis or not support your hypothesis
    • If you have a "finding" that holds only for some regions (e.g., "Africa"; "Old World regions = 1,2,3") but not for others you may want to ask yourself: what is it about these specific regions that might make the hypothesis work there but not elsewhere? For example, the "Old World" regions (1,2,3) all have large domesticated animals, the new world had none (except the llama). Bridewealth exchanges at marriage involving payment of the brides family in large animals consistute a large-scale marriage "system" in parts of the old world, but not the new, with very different implications. Also very different effects on disease and depopulation when the "Old World" societies invaded those of the "New World."
    • WHEN YOU DONT KNOW WHAT SOMETHING MEANS IN THE CODEBOOK (e.g. "bridewealth") you can email me to ask (or ask in class)
    • Sometimes your hypothesis doesnt work worldwide (no overall correlation) but it works in certain regions. That is a "lead" to follow in your research.
    • This kind of "third factor control" can also be used, as your project progresses, to control for other variables besides region, and test possible conditions under which a hypothesis works (improved correlation) or doesnt work (diminished correlation).

    3. The class seemed to go pretty well Tuesday: review the 4th point from the last assignment on Organizing your time - working together. Lots of people are helping one another and cooperating to get their research done and figure out how to use the computers as well as the library. Dont miss out on being part of the cooperative web in the class! Pauline Manaka at the library also also been a great help and we will find a time for the whole class to go over and get an introduction to finding library resources for your topics/projects. And many students have emailed me; I have been able to respond to most and just responded to a couple of others (Beth, Aghdas: see your email). I plan to be healthy after this flu and back again on TUESDAY Feb 2nd, available on office hours that day 1-2PM 4169SSPA, and available WEDNESDAY much of the day; then to get you feedback on your proposals if possible by Thursday.

    4. IN GETTING YOURSELF ORGANIZED:

    • Why not put in a ring binder all the materials you need to bring to the lab each Tues, Thurs to get your work done:
    • a. The one page with numbers of Spss files so you can locate variables
    • b. The Spss codebook by Divale
    • c. Your MAPTAB Tutorial and Users Manual (they are stapled together)
    • d. Your notes on topics, variable numbers you are using
    • e. A floppy disk to copy files you want to save
    • f. The list of topics and variables (from 3rd week: remember the numbers in the first columns are merely to sort the topical organization; the variable numbers are in the second column and are followed by a period and a lower-case variable name plus the correct variable number.
    • g. Printouts that you should make of THIS PAGE and the Jan 24th Assignment page -- containing important instructions
    • A research and writing class is largely ABOUT GETTING YOURSELF ORGANIZED to be a bit creative in your research project. Dont sit back passively!



    1. (no longer used in Anthro 174AW) If you use MAPTAB:

    • Hit Return until you get to "Select variables by ..." with the 5 numbered options
    • Enter option 4: "Select variables by Variable Number"
    • Enter the numbers 200, ..., ... -- three numbers, separated by commas, for the variables you want, plus any additional variables you want as well
    • After the files are read, the main menu will appear and you choose option 6 Tabulate and then 1 Cross-Tabs
    • If you want to SAVE RESULTING TABLE TO DISK, enter a file name, otherwise press RETURN or ENTER
    • For COLUMN VARIABLE enter your first variable's number
    • For ROW VARIABLE enter your second variable's number
    • For CONTROL VARIABLE enter the number 200
    • keep hitting RETURN until you get the table
    • To print your outfile you need to return to the main menu of MAPTAB, go to START, open ACCESSORIES, open the wordprocessor, and in the wordprocessor open the filename you gave earlier to save the table. You can also copy the file you save to disk drive a: if you insert your own floppy disk.