Site Map

Problem Sets

Week 7 Project:

An ancient mummy has been recovered in Peru. We are interested in knowing something about the history of the population from which this individual comes. Having mastered discriminant function analyses, you are able to say something about how this individual compares to other populations. You choose to include comparative samples from Japan, Alaska, North America (Indian Knoll), and Roman Britain. You are interested in predicting group membership on the basis of body proportions. Your collaborator, Ben, has sent you measurements from these populations for humeral length, femoral length, and bi-iliac breadth. Based on cranial and pelvic features, you determine that your Peruvian individual is female. You have therefore restricted your comparative samples to females. The data file contains measures for the four comparative samples and the Peruvian individual.

Address the following:

  • Is there an effect of group membership on body proportions among your four comparative samples? Write one sentence including evidence to support your conclusion.

  • To what group is your Peruvian individual most likely to belong based on body proportions? Present a graph to demonstrate this relationship, and include a statement of relative probabilities.

  • Describe how this individual differs from the group means for the comparative samples in terms of humeral, femoral, and bi-iliac measures (i.e., discuss loadings of dependent variables on the discriminant functions).

  • DATA FILE: (right click and choose "save as...")
    [NOTE: This is an Excel file, which will need to be imported into SPSS.]

    Week 7.xls

    Week 3 Project:

    You were fascinated to learn in the Monday stats class (among other things) that breadth of the knee joint appears to be more highly correlated with body mass than locomotor behavior. Knowing that bi-iliac breadth is also used to estimate body mass, you decide to investigate the relationship between the knee joint and pelvic breadth. Using left tibial plateau breadth (LTPB) and bi-iliac breadth (BIB) from Ben's Goldman dataset, conduct a regression analysis on the relationship between these two variables. Use LTPB as your independent variable and BIB as your dependent variable.

  • Remember to address assumptions about the distribution of your variables in your sample.

  • Write the equations for the LS and RMA axes, and think about why you would choose to use one or the other.

  • Prepare a graph for publication showing the LS regression model. Think about what you might want to convey to an audience, and include relevant details in the graph (e.g., confidence intervals, labeling outliers).

  • Using your data, write a one sentence description of your regression results.

  • Determine the mean LTPB value in your sample, and using your regression model, provide a 95% confidence interval for the predicted value of BIB given that LTPB.

  • DATA FILES: (right click and choose "save as...")
    [NOTE: These are Excel files, which will need to be imported into SPSS.]

    Week3 A-team.xls (Indian Knoll)
    Week3 B-team.xls (Germans)
    Week3 C-team.xls (Point Hope)

    Week 2 Project:

    Ben has returned from distant travels and sent you properly formatted data. You want to know if humeral length varies based on geographic origin.

    Team A: You want to test the hypothesis that Egyptian individuals tend to have shorter arms than Europeans.
    Team B: You want to know if humeral length varies across Native American groups.
    Team C: You want to test the hypothesis that humeral length varies with possible migration routes around the Pacific Ocean.
  • Choose two levels (values) of your independent variable ("location") from your dataset and compare them using the t statistic. Report findings.

  • Compare all three levels of the independent variable using an ANOVA. Consider your assumptions, and prepare any planned contrasts or post hoc comparisons to test or explain your hypotheses/results. Prepare brief output (in .OUT or .DOC format) to bring to class on Thursday, including graphs where they are useful and a statement summarizing/interpreting the results.
  • DATA FILES: (right click and choose "save as...")
    [NOTE: These are Excel files, which will need to be imported into SPSS.]

    Week2 A-team.xls
    Week2 B-team.xls
    Week2 C-team.xls

    Week 1 Project:

    Imagine that you and your collaborator (aka Ben) are writing a paper about the skeletal biology of a museum sample. Ben has collected the maximum length of the left humerus (LHML), and prior to departing for parts unknown, he has sent you this data file.

  • Get the data into SPSS (or other statistical package) and format your variables properly. Save the .sav file.

  • Calculate descriptive statistics for this sample (one sample, ignore distinctions in "Note" field if present), including measures of central tendency, dispersion, and normality. Save the .out file.

  • Summarize relevant information as if for publication. Think about how you can best convey relevant information, without just listing a bunch of statistics. Include data in table and graphic formats. Save this as a .doc file. Consider the following:
  • a) Is the mean LHML the best measure?
    b) Do you have any outliers? If so, how could you report this information?
  • You have developed an interest in studying sexual dimorphism of the LHML value in this sample. Sex is coded with 0 for males and 1 for females. Use boxplots, histograms, or other graphics to visualize the presence/absence of this difference. Save this as a .doc file.
  • Print and bring your .doc files to discussion on Thursday. I'll have the data files available for discussion. DON'T SPEND MORE THAN AN HOUR ON GETTING YOUR DATA FORMATTED AND ANALYZED. If you have trouble, or if you have any questions, come talk to me.

    DATA FILES: (right click and choose "save as...")

    Week1 A-team.txt
    Week1 B-team.xls
    Week1 C-team.txt

    Return to Valerie DeLeon's page.