Assignments


COM 508 Quantitative Foundations of Communication
PSC 508 Basic Statistics for the Social Sciences

Dr. Frank Tutzauer (comfrank@acsu.buffalo.edu)
Fall 1999 MWF 12:00 noon - 12:50 pm (note revised time)

Preliminaries:

Stylistic Rules for Completing Homework

Assignment #1:

If you wish, you may download the Class Example Data File that we have been using. The data concern a hypothetical medical experiment, and each line of data contains information concerning the results of a drug therapy administered to a number of sick patients. The format of the file is as follows:

Columns

Variable Comment

1-3

ID ID number

5

Treatment 1 = received drugs
2 = control group

7

Result 0 = patient died
1 = patient lived

8

Personality 0 = Type A
1 = Type B

10-12

Psychosis rating  

13-15

Stress rating  

16-18

Sexual deviance rating  

The three ratings are interval-level variables. Higher numbers mean more psychosis, stress, or sexual deviance, respectively.

You may download the data file in either text format or as an SPSS .sav file. Note: The SPSS .sav file was created using the latest version of SPSS for Windows. You may have problems opening this file with earlier versions of SPSS. If so, let me know and I will also put up a version in SPSS portable format, which theoretically should be compatible with ALL versions of SPSS (even mainframe).


Some Notes about the Assignments:


On any assignment that requires computer output, hand in to me the following: your SPSS syntax, your output, and any answers to specific questions. When you are asked to test hypotheses, be specific--for example, "We reject the null hypothesis that the mean is 15," or "The main effect shows that drug treatment A contributed to psychosis."


This next batch of assignments deals with distributions, particularly the binomial. Assignment 2 (
assign2.txt) asks you to generate some histograms. In Assignment 3 (assign3.doc or assign3.txt) you use the binomial distribution to test some hypotheses about proportions, and Assignment 4 (assign4.doc or assign4.txt) lets you explore the normal approximation to the binomial.


Midterm Make Up Problems:

Download the make up problems in either Word format or text format. You will also need to download an SPSS sav file. If you are eligible to do make up problems, then you will find on the front of your exam one or more numbers written in green. These numbers are the problems that you will be allowed to make up. They correspond to the numbers on the make up problems that you just downloaded. If you are doing make up problems, you must do all of the ones listed in green (i.e., you cannot pick and choose). Answer all of the indicated make problem(s). Be complete. Be neat. Be thorough. Show me that you've mastered the material. Whenever a hypothesis test is asked for, be sure to include an explicit statement as to what your conclusion is. For example, "Based on the t-test, t(45) = 3.51, p < .05, one-tailed, we conclude that Group A was significantly smarter than Group B." When you turn in your answers, turn in your original exam as well. We'll set a due date later.


Practice Exam:


This practice exam concerns some hypothetical data about....beer! You can download the
text data file, the SPSS sav file, and the practice exam questions. This exam is very similar in style to what will be asked on the final.


Final Exam:


The final exam is due Wednesday, December 15. You can leave it in my mailbox in Baldy 356, or slip it under my door at Baldy 361. The exam is open notes and open book, but you should not consult other people. You should download the
questions and the SPSS sav file. There is no text data file. Load the sav file into SPSS and answer the indicated questions. For each question, write out a short answer based on the SPSS output. In the answer, you should give an explicit statement as to what effects there are, if any. In other words, don't just say "Joint authority is related to settlement." Instead, say something like "Joint Authority is related to settlement--the high authority bargainers achieved significantly higher settlements that the low authority bargainers" (or whatever specific results you find). In all cases, report IN THE ANSWERS, whatever statistics, degrees of freedom, p values, etc., you used to form your conclusions. In all cases, make sure to check the validity of relevant assumptions.


What do I want from you? Your SPSS output along with your answers, typed and keyed to the output. An SPSS syntax file is optional.


Have fun!


Back to the course homepage