Mdde+602+Assignment+2

**Assignment 2:** //Analyzing Quantitative Data//

 * Credit weight:** 20%
 * Due date:** End of week 9
 * Readings:** Neuman Chapter 12; SPSS 16.0 for windows; MDDE602datamod.sav (SPSSSV data file)

Goal of the assignment
The quantitative analysis of data has been the mainstay of both the deductive and inductive sciences for centuries. In particular, the advancements made the mathematicians and logicians of the 19th century have provided researchers with the theoretical foundation for the probability theorems and associated statistical tests that have allowed researchers to control for the effects of extraneous factors and chance. Associated with these developments has been the evolution of tools to facilitate the various analyses, ranging from mechanical computational devices to the digital computers. With respect to the latter, programs have been created that greatly reduce the time-consuming tasks of recording and analyzing the data that have resulted from experiments and other types of investigations. In the social sciences, a number of these programs have become a requisite for accurate and efficient data analysis. One of the most popular, the Statistical Program for the Social Sciences, has been regarded as an indispensable component of high quality research for many decades. Graduate students in the social sciences should have at least a basic knowledge of how to use such a computer-based tool in order to appreciate fully how the reporters of research have used such tools to investigate problems or questions of interest. Assignment 2 focuses on the acquisition of skills that are associated with the use of such a commonly used data analysis tool.

SPSS Student Version
 SPSS (Statistical Package for the Social Sciences) is a very powerful set of application programs for analyzing numerical data. SPSS can take data from various file types (e.g. Word files and tables, Excel spreadsheet files and Access database files) and use them to generate tabulated reports, charts, descriptive statistics and complex statistical analyses. Alternatively, data can be entered via the data editor, a spreadsheet-like application that is part of SPSS. The SPSS Student Version (SPSSSV) is a scaled-down version of the full SPSS package yet it is fully featured and allows for all but the most complex and large-scale analyses to be run. None required as the software is on CD-ROM and is included in the course materials package. To install SPSSSV for Windows, simply insert the CD-ROM into your CD-ROM player, select the "Install SPSS Student Version" from the opening screen and respond to the prompts. Then follow directions to load and license your program. Your Student Version license allows you to install the program only twice. If you attempt installaion more than twice you will need to call SPSS tech support for help. Allow plenty of time to be on hold. If your CD-ROM player is not set to "auto-run" - i.e., to open the root directory of the disk automatically and immediately after it has been inserted, then you will have to use the Windows Explorer program to access the "Setup.EXE" file in the root directory on the CD-ROM. In this instance, navigate via the Explorer to your CD-ROM player (usually drive D: but could be E:, F: or even G: if you have a number of additional peripherals installed - e.g., a second hard drive, a Zip drive), double-click on the "Setup.EXE" file and, as above, select the "Install SPSS Student Version" from the opening screen. You can ignore the "Browse the CD-ROM" option. Note that it is recommended that you accept the default settings during the installation process. For example, the set-up program will prompt you to save the SPSSSV files in the "c:\Program Files\SPSS" folder/directory on your hard drive. You have the option of saving the files to another folder/sub-directory but it is recommended that you not do so.
 * Description:**
 * Downloading:**
 * Installation:**

General comments re SPSSSV
> > a. "My Documents\SPSS\SPSSSV output\working > b. "My Documents\SPSS\SPSSSV output\final > > You might also wish to create a folder/sub-directory for your course papers.
 * 1) Data for the analysis for assignment is in MDDE602datamod.sav which was sent to you by e-mail or is posted in the Assignments Forum.
 * 2) Make a backup of the "MDDE602datamod.sav" file immediately as it is possible (but not likely) that this file will become corrupted as a result of, for example, a power failure, or nefarious gremlins clobbering the file. You may wish to rename the file "MDDE602datamod.BAK" and store it in another folder/sub-directory or on an external storage medium (e.g., ZIP disk, floppy disk, CD-ROM). You may also request another copy from the professor.
 * 3) The data file "MDDE602datamod.sav" that is provided to you contains 11 variables in 11 columns. The labels at the top of each column indicate the variable name and the data type (nominal, ordinal, interval). There are values for 30 students/subjects.
 * 4) Create a series of folders/sub-directories on your hard drive in which you will store the results of your efforts during the course. A suggestion is that you create two folders/sub-directories to store your SPSSSV output: a "working" folder/sub-directory to store your first attempts to generate output and a "final" folder/sub-directory to store your polished version.
 * 1) Note that the extension for all SPSSSV output files is ".SPV" for version 16.0 and above. This extension will be added automatically to any output file that you save. Note that you should use the filenaming convention (see "What to turn in" section below) when you save the output for each SPSSSV question.
 * 2) It will be assumed that, before you attempt to do the assignment, you will have worked through the tutorial on the CD-ROM and that you have at least skimmed relevant chapters in the SPSSSV manual.
 * 3) Note that you can edit the SPSSV output files that result from your statistical analyses. The leftmost column allows for deleting the results of an analysis and also for moving each set of the items up or down the list of output items.
 * 4) If you opt to print the contents of the output file, the graphs that appear red on your computer monitor will be printed black (unless you have a color printer) and will consume a great deal of toner. It is not necessary to print the contents of your SPSSSV files to meet the submission requirements for Assignment 2.

SPSSSV Questions
A number of points need to be taken into consideration before responding to the eight questions below: > For all procedures described below, the SPSSSV data editor must be loaded with the contents of the "MDDE602datamod.sav" file.When you launch SPSS you will see a window with a list of options for you to do. About half way down the window is an option to open an existing data source. Click on that. You'll then see an window that looks like My Documents. This window is a menu where you can locate the data module that you saved. Click on the data module file and it will open. You'll see a spreadsheet with names, classes, and scores. This is the data you will use for your assignment. > > From this point, you just need to follow the instructions step by step and you'll be able to generate the charts you need to complete the problems. Use Neuman, the SPSS manual (really good), and any other resources you have to understand the problems. I already know the answers so you just need to convince me that you understand. > Most questions focus on the description of the characteristics of a set of data - mean, standard deviation, correlation coefficients - and the results do not allow for any generalizations to hypothetical populations of individuals who are similar to those who generated the original scores - e.g., other MDDE 601, MDDE 602 or MDDE 603 students. In other words, no inferences or generalizations can be made from the results of the statistical analyses. Question 6 analysis does lend itself to generalization - i.e., inferences can be made. > Each question below has two components. The first, "SPSSSV procedures", provides step-by-step directions on how to get SPSSSV to carry out the analysis that is required. The second, "Question #", requires a written response that you will submit in a Word file that accompanies the SPSSSV output files.
 * 1) //**How do I open the data module?**//
 * 1) //**Do the data reflect real scores and real students?**//
 * 1) //**What does each question include?**//

Assignment questions
****Question 1:** Although there are only 30 scores for the variable "i601" (MDDE 601 Final Grade Interval), does the distribution of scores for this variable appear to reflect a normal distribution?** **Please explain how you determine whether or not the scores reflect a normal distribution.** Review output from question one and answer the following question: (see additional information at: **http://www.surveysystem.com/signif.htm**) //Note:// SPSSSV does not generate values for both the median and the mode for the "Statistics/Descriptives" option. (see additional information at: **http://www.wellesley.edu/Psychology/Psych205/pairttest.html** and Neuman pp.370-373). For this question, we will assume that our 30 students participated in an experimental study and were randomly selected from the population of all MDDE program students. Thus, we can make inferences (generalizations) from our sample of 30 subjects to the larger population (i.e., all MDDE program students). So, forget that the perspective that we have had about them until now. That is, they should not be viewed as students who have been registered in 3 MDDE courses but rather should be individuals who have agreed to act as subjects in a study.
 * 1. Frequency distributions** (Neuman p. 346)
 * SPSSSV procedures:**
 * Go to the Data editor
 * Select Analyze, Descriptive, Frequencies
 * Move "MDDE 601 Final Grade Interval into the "Variable" box
 * Click statistics
 * Central tendency: check mean, median, mode; Distribution: check skewness and kurtosis
 * Click continue, then charts
 * Check histograms and with normal curve option
 * Click continue
 * Click 'OK'
 * Double-click on the 'Frequencies' title box on the right-side of the screen and replace the label with 'Question 1'
 * Save the output file according to the preferred naming convention described in the "What to submit" section below.
 * 2. Bar chart (pp. 348-349) **
 * //Question 2://**** **How does the bar graph for this output differ from the histogram in question 1? Does this bar graph reflect a normal distribution? Is the type of presentation of data important in the communication of findings? Please explain your answers.**
 * SPSSSV procedures:**
 * Go to the Data editor
 * Select Analyze, Descriptive, Frequencies
 * Move MDDE 601 Final Grade Interval into the "Variable" box
 * Click statistics
 * Central tendency: check mean, median, mode
 * Click continue, then charts
 * Check barchart and percentages
 * Click continue
 * Click "OK"
 * Double-click on the "Frequencies" title box on the right-side of the screen and replace the label with "Question 2"
 * Save the output file according to the preferred naming convention described in the "What to submit" section below.
 * 3. Central tendency** (pp. 349-350)
 * Question 3:** Explain how the values for the mean, median and mode for the continuous variable MDDE 601 Final Grade Interval relate to each other for this set of scores. For this distribution is the mean, median or mode, the best measure of central tendancy? Explain your answer.
 * 4. Central tendency for nominal variables** (pp. 349-350)
 * //Question 4://**** **Report the central tendency of the variable gender. Explain how values for the mean, median and mode should be used in interpreting the central tendency of nominal variables such as gender.**
 * SPSSSV procedures:**
 * Go to the Data editor
 * Select Analyze, Descriptive, Frequencies
 * Move "Gender" into the "Variable" box
 * Click statistics
 * Central tendency: check mean, median and mode
 * Click continue, then charts
 * Check barchart and percentages
 * Click continue
 * Click 'OK'
 * Double-click on the "Frequencies" title box on the right-side of the screen and replace the label with "Question 4"
 * Save the output file according to the preferred naming convention described in the "What to submit" section below.
 * 5. Comparisons across groups** (Neuman p. 351)
 * //Question 5(a)://** What is the relationship between the mean values and the standard deviation values for each of the 3 variables? Explain your answer. Explain homogeneity and heterogeneity across these distributions.**
 * //Question 5(b)://** Can one determine whether or not the mean values for each of the 3 variables are significantly different from each other as a result of this analysis?**
 * SPSSSV procedures:**
 * Go to the Data editor
 * Select "Analyze/Descriptive Statistics/Descriptives"
 * Ensure that "MDDE 601 Final Grade Interval" is entered into the "Variable(s)" box
 * Ensure that "MDDE 602 Final Grade Interval" is entered into the "Variable(s)" box
 * Ensure that "MDDE 603 Final Grade Interval" is entered into the "Variable(s)" box
 * Click the "Options" option
 * Select the "Mean, Std. Deviation, Variance, Minimum, Maximum" options. Deselect the other options if necessary
 * Select "Continue"
 * Click on 'OK' in the 'Descriptives' window
 * Double-click on the 'Descriptives' box on the right-side of the screen and replace the label 'Descriptives' with "Question 5"
 * Save the output file according to the preferred naming convention described in the "What to submit" section below.
 * 6. T-test: correlated/paired samples**
 * //Question 6://**** **Is there a significant difference for the two conferencing systems? Do we accept or reject the null hypothesis? Discuss your reasoning.**

The purpose of the study was to determine the effect, if any, of two conferencing systems (WWWBoard, WebCT) on the number and quality of the messages that are posted (a dubious hypothesis but we'll ignore that for now). The 30 subjects were exposed to two levels of an independent variable (e.g., level 1 - the use of the WWWBoard conferencing software and level 2 - the use of the WebCT conferencing software). The scores - the dependent variable - that result from the use of the conferencing systems are recorded in the two variables "itrial1" (WWWBoard) and "itrial2" (WebCT). The scores are the marks awarded to each subject for the number and quality of the messages posted while using WWWBoard and WebCT.

Note 1: Both "itrial1" and "itrial2" are instances of an interval data type. Note 2: The alpha level has been set at .05 (i.e., p < .05).

The problem or question for this study, then, is the following: What is the effect of both conferencing systems on the number and quality of the posted messages and, if there is a differential effect, is it significant - i.e., beyond chance? The null hypothesis is the following: There is no significant difference in the effect of the two conferencing systems on the number and quality of the posted messages.

There are two steps for this analysis. The first is to determine if the data for the dependent variable - "itrial1" and "itrial2" - meet certain assumptions. The most important assumption is that the data are normally distributed. If this is true, then the second step is to execute an appropriate statistical test to determine if the means are significantly different. Because the data - i.e., the scores for the variables "itrial1" and "itrial2" - were generated by the same sample, the correct statistical test is the correlated or paired samples t-test.

// To determine whether or not a statistical test is significant (e.g., two means have been drawn from two different populations, the correlation between two variables is beyond chance) or not (e.g., two means have been drawn from the same population, the correlation between two variables does not exceed chance), the pertinent part of the SPSSSV output is the following table. Specifically, the decimal number under the heading "Sig" (.000 in the table below) is the significance level or criterion level that informs the researcher as to whether or not the data-based calculated value (-7.34) is significantly different or not from an ideal or theoretical distribution (i.e., the population from which the sample was drawn). If this significance value were less than the alpha level that was preset by the research (e.g., p < .05), then the conclusion would be that the null hypothesis is rejected. If the significance value (e.g., .078) were equal to or greater than the alpha level, then the null hypothesis would not be rejected .// ||
 * ===//Hint://===
 * t || df || Sig. (2-tailed) ||
 * -7.34 || 48 || .000 || ||

First step (Are the data are normally distributed?) Variable "itrial1" Variable "itrial2" Second step (Paired samples t-test) //Note:// The two variables - i602, i603 - are interval data types. Generate a scattergram for the relationship between "age" and "MDDE 601 Final Grade Interval". Generate a scattergram for the relationship between "age" and "MDDE 602 Final Grade Interval". Generate a scattergram for the relationship between "age" and "MDDE 603 Final Grade Interval."  (Neuman, pp. 363-365)
 * SPSSSV procedures:**
 * Go to the Data editor
 * Select "Graphs/Legacy Dialogs/Histogram"
 * Select "Trial 1 - Interval" and enter into the "Variable" box
 * Check the "Display normal curve" option
 * Click 'OK'
 * Double-click on the "Graph" box on the right-side of the screen and replace the label "Graph" with "Question 6"
 * Go to the Data editor
 * Select "Graphs/Legacy Dialogs/Histogram"
 * Select "Trial 2 - Interval" and enter into the "Variable" box
 * Check the "Display normal curve" option
 * Click 'OK'
 * Note:** The data for both variables should approximate a normal distribution so the application of the paired samples t-test is warranted.
 * Go to the Data editor
 * Select "Analyze/Compare Means/Paired-samples T Test"
 * Select "Trial 1 - Interval" and "Trial 2 - Interval" and enter into the "Paired variables" box. Note that you must select the two variables from the leftmost column
 * Click 'OK'
 * Save the output file according to the preferred naming convention described in the "What to submit" section below.
 * 7. Bivariate relationships** (Neuman pp. 354-356)
 * //Question 7://**** **Do any of the three scattergrams reflect a moderate or strong linear relationship? Discuss the strengths and relationships presented by the three scattergrams.**
 * SPSSSV procedures:**
 * Go to the Data editor
 * Select "Graphs/Legacy Dialogs/ScatterDot"
 * Select the "Simple Scatter" option and click on "Define"
 * Select "age" and enter into the "Y-axis" box
 * Select "MDDE 601 Final Grade Interval" and enter into the "X-axis" box
 * Click 'OK'
 * Double-click on the "Graph" box on the right-side of the screen and replace the label "Graph" with "Question 7"
 * Follow the same steps as for "age" and "MDDE 601 Final Grade Interval."
 * Select "age" and move into the "Y-axis" box
 * Select "MDDE 602 Final Grade Interval" and move into the "X-axis" box
 * Click 'OK'
 * There is no need to edit the title "Graph" for this analysis
 * Follow the same steps as for "age" and "MDDE 601 Final Grade Interval."
 * Select "age" and move into the "Y-axis" box
 * Select "MDDE 603 Final Grade Interval" and move into the "X-axis" box
 * Click 'OK'
 * There is no need to edit the title "Graph" for this analysis
 * Save the output file according to the preferred naming convention described in the "What to submit" section below.
 * 8. Correlation coefficient/Correlation matrix**
 * //Question 8(a)://** Which of the correlations are significant? Explain how this was determined. If any, how should such a significant correlation be interpreted? What are possible explanations for the relationship presented in this analysis?

//**Question 8(b):**// Why are the values of 4 of the 16 correlation coefficients 1.000? Please explain.** //Note:// The three variables - i601, i602, i603 - are interval data types. Consequently, the Pearson's Product-moment correlation procedure is appropriate.
 * SPSSSV procedures:**
 * Go to the Data editor
 * Select "Analyze/Correlate/Bivariate"
 * Ensure that "MDDE 601 Final Grade Interval" is entered into the "Variables" box
 * Ensure that "MDDE 602 Final Grade Interval" is entered into the "Variables" box
 * Ensure that "MDDE 603 Final Grade Interval" is entered into the "Variables" box
 * Ensure that "age" is entered into the "Variables" box
 * Ensure that only the "Pearson" option is selected in the "Correlation coefficients" box
 * Ensure that the "Two-tailed" option is selected in the "Test of Significance" box
 * Click 'OK'
 * Double-click on the "Correlations" box on the right-side of the screen and replace the label "Correlations" with "Question 8"
 * Save the output file according to the preferred naming convention described in the "What to submit" section below.
 * 9. Spearman's Rho** (Neuman, pp. 362-364)
 * //Question 9(a)://** Is the correlation between the variable "orank ("Admission Rank Order - Ordinal") and the variable "i601" ("MDDE 601 Final Grades - Interval") significant? Please explain.

//Note:// The variable, "orank", is an instance of an ordinal data type and "i601" is an instance of an interval data type.
 * //Question 9(b)://** Why was the Spearman's rho procedure used to determine the relationship between the two variables?**
 * SPSSSV procedures:**
 * Go to the Data editor
 * Select "Analyze/Correlate/Bivariate"
 * Ensure that "Admission Rank Order - Ordinal" entered into the "Variables" box
 * Ensure that "MDDE 601 Final Grade Interval" is entered into the "Variables" box
 * Ensure that only the "Spearman" option is selected in the "Correlation coefficients" box
 * Ensure that the "Two-tailed" option is selected in the "Test of Significance" box
 * Click 'OK'
 * Double-click on the "Nonparametric Correlations" box on the right-side of the screen and replace the label "Nonparametric Correlations" with "Question 9"
 * Save the output file according to the preferred naming convention described in the "What to submit" section below.

What to submit:
> > km_1.SPV > km_2.SPV > km_4.SPV > km_5.SPV > km_6.SPV > km_7.SPV > km_8.SPV > km_9.SPV > > //Note:// Insert a page break (in Microsoft Word - select "Insert/Break/Page break") at the end of the title page.
 * 1) The nine files to be submitted for Assignment 2 are listed below: eight SPSS files and one Word File. SPSSSV output filenames for each question include the initial from your last name, followed by the initial from your first name, followed by an underscored, then the question number. For example for student Marguerite Koole, the file names would be:
 * 1) **A Microsoft Word file** that includes your full answer to the question (the "show your work" part of each question). Each answer should be half to one full page. Include a title with designation, name, student ID number, and assignment number. Check APA style format to see how to arrange the cover page.

Course Author: Martha Cleveland-Innes, Ph.D. Production/Research Editor: Mickey MacGregor
 * The Course Team**