Quercus Project Manual and Teaching Notes

QUERCUS

Version 3.1

A Teaching and Learning Technology Programme Project funded jointly by the four UK higher education funding bodies, HEFCE, HEFCW, SHEFC and DENI

Quercus v3.1 Copyright (c) 1996 Department of Statistics and Modelling Science, Strathclyde University, Glasgow G1 1XH, Scotland.

Download this page as a Word for Windows v6 document (2.5 Mbytes), or in ZIP format (175 Kbytes).

Contents

Welcome to Quercus!

Quercus is a collection of courseware modules aimed at helping Bioscience students learn basic statistics. It was developed using Scottish Higher Education Funding Council funding under the Teaching and Learning Technology Programme by a consortium of universities including Strathclyde, Heriot-Watt, Edinburgh, and Stirling Universities. Quercus is available on both Mac and PC platforms.

Quercus consists of two types of module, plus a variety of ancillary and resource materials, intended to cover the material of an introductory Statistics course which addresses the needs of undergraduate Bioscience students. In addition to covering statistics topics, Quercus includes material on how to use its workbooks, provides an introduction to the statistical package Minitab for Windows (TM) v9 or higher, and covers the writing of reports.

WorkBooks: There are ten workbooks (green icons) in Quercus; each isintended to be run in a one-hour supervised computer laboratory session, and covers a specific topic.

Exercise Books:There is an Exercise Book (blue icon) for each WorkBook (except WorkBook 10), providing extra practice with the material and techniques covered, and a short self-assessment test for students and lecturers to monitor progress and problems. The Exercise Books are intended to be run in less than one hour's unsupervised laboratory time.

If you think of Quercus as a "book", then the WorkBooks are roughly equivalent to chapters, and the Exercise Books are computer-based problems and tests associated with those chapters.A given course might not use all of the workbooks and exercise books, or deal with them in a different order than they are presented here.

Part 1:

Installation Guide for Version 3.1

1. Introduction

The installation program INSTALL.EXE copies the files needed for using the Quercus collection of Computer Aided Learning applications from one or more floppies to a nominated directory on a hard drive.

In addition to making sure that all files get to the right place, it also has to put correct path information into a variety of INI and TXT files which will be used by the Quercus workbooks and ancillary programs, and create a Windows Program Manager Quercus group. These ends are achieved by using an INI file for INSTALL.EXE to provide fixed installation information, and through obtaining information from the user during installation. These aspects will be considered in turn in this document.

1.1 Installing Quercus

1. The program INSTALL.EXE must be run from a floppy (the first in the installation set of floppies), as it will not permit installation onto the drive from which it was launched.

2. When Install starts, the first thing it does is to establish communications with the Program Manager. As a result, the Program Manager is activated (really noticeable only if the system uses an alternative desktop scheme).

3. When the Install window appears, it has two pull down menu items: Options and Help. Under Help there is the single selection About, causing program information to appear. Under Options there are two selections: Install and Exit. Selecting Install starts the installation process. See the diagram below in Fig 1.

   Figure 1: The Install Programme Screen.

   

4. The INI file is read, including details of all the applications to appear in the Quercus group.

5. The user is prompted for a QUERCUS directory, on the drive where Windows is installed. Cancellation ends the program. A choice causes Install to create the QUERCUS, QFILES and MFILES directories if necessary. Note: the QUERCUS path must end in a \ character.

6. The user is prompted for a path for MINITAB, starting at the root directory of the target drive. The validity of this result is not tested, but as selection comes from a standard file dialogue, the chance of error can be reduced. An error can be dealt with later by editing the workbook config (CFG) files.

7. The content of each disk in turn is copied from the floppy directories to their counter-parts on the hard disk. The user is prompted when each new disk is required. In the current version there is no other progress indication given.

8. The stored details of the Quercus group items is now updated with the appropriate paths, and DDE communication with the Program Manager produces a group called "Quercus" containing all the specified items. If there was a previous group of that name, it is deleted not added to.

9. The contents of all CFG files in the QFILES directory are updated to conform to the installed directories and paths (see CFG Files below).

10. The user is told that installation is finished. Use of the Action|Exit menu selection ends Install.

1.2 Directories

When INSTALL.EXE installs the Quercus collection, it does so in a directory on the nominated hard disk (hereafter the "Quercus directory"). Within that directory there will be created two more directories: one called QFILES, and one called MFILES. By default, the MFILES directory contains only Minitab *.MTW data files. The QFILES directory contains all the resource and configuration files for the workbooks, DLLs, and the workbooks themselves. The files in the QUERCUS directory itself will include utility programs and their INI files, and most importantly the AuthorWare Professional distributable application package, RUNAPW2.EXE.

Each installation floppy must be organised in the following manner: there must be three sub-directories in the main directory of the floppy, called QUERCUS, QFILES and MFILES respectively. Obviously, the distribution of each of the kinds of files across the installation disks is immaterial to their final disposition, as all files in any disk QUERCUS directory end up in the hard disk QUERCUS directory, and all files in any disk QFILES and MFILES directories end up in the respective subdirectories of the QUERCUS directory. The choice of which files go on what disk will depend on minimising the number of disks to be installed, offset against the convenience of keeping all the files for a given workbook on the same floppy disk.

1.3 INSTALL.INI

All text in the INSTALL.INI INI file is case-insensitive. The [install] section contains a item disks, an integer which tells Install.exe how many floppies are to be used during installation. As each is copied in turn, the user is prompted for the next, until disks have been copied. In the following example there are two disks in the installation set:

[install]
disks=3

In this version of INSTALL.EXE, there is another item under [install], linecfgs, an integer valued switch telling Install whether the old style line-by-line CFG files are to be used, or the new INI style CFG format. Setting linecfgs to 1 causes line-by-line format to be assumed; setting it to 0, or leaving it out, causes INI style format to be assumed. In later versions this feature will disappear, as and when we move completely to the INI style CFGs.

Once all the files on the installation disks have been copied, Install.exe makes use of the rest of the INI information to create a Program Manager group called "Quercus". Each program in the group is specified by an INI section with the following format:

[quercus ]
filename=...
runby=...
title=...
iconnumber=...

filename
Each section is defined by a letter (a, b, ..., z), which must start at a and have no gaps. Within the section, the item filename gives the "partial path" of the program in question; if a program called Ortho.app is to be found in the QFILES subdirectory, then we would have filename=qfiles\Ortho.app.

runby
For some applications (in particular, the AuthorWare applications used by Quercus) it is necessary to specify the path of a program which will cause them to run. The runby item gives the partial path of the running application. For example, if RUNAPW2.EXE in the QUERCUS directory is to be used, then we have runby =runapw2.exe.

title
Items in a Program Manager group have a caption below them, and this is provided by the title item in each section. For example, if an application relates to data analysis, we might have title=Data Analysis.

iconnumber
For most executable Windows files, an icon will usually be provided, and this will automatically be used by Program Manager. If this icon is not appropriate or sufficiently informative, you can choose another within Program Manager. The iconnumber item allows this to be done at installation. The integer number referred to by this item is the index of the icons provided in the file QUERCUS.DLL; this index starts at 0, and at present goes to 4. For example if we wanted icon number 2, we simply have iconnumber=2.

In the following example, the application is the fourth (letter d!) in the list of those to go into the Quercus group. Its file is called WRKBK1.APP, which is in the QFILES directory. The program RUNAPW2.EXE is need to run it, and it makes use of icon number 2 in QUERCUS.DLL:

[quercus d]
filename=qfiles\Wrkbk1.app
runby=Runapw2.exe
title=Exploring data
iconnumber=2

1.4 CFG FILES

These are text files used by the Quercus workbooks to provide user-adjustable access to text resources, external packages (e.g. Minitab), and utility programs. Once installed, each line of a CFG file (the format is .cfg) contains exactly one file path. In preparing CFG files for distribution with Install, the following guidelines should be adhered to (note: there are no quotation marks or spaces; even trailing spaces):

(a) Where reference to MINITAB is intended, make the line contain only the text:
%%MINITAB%%

In particular watch out for trailing white space. Install will replace this with the user's chosen path for Minitab.

(b) For applications which reside in the Windows directory (e.g. NOTEPAD.EXE, WRITE.EXE), use the text:

%%WINDOWS%%

e.g.
%%WINDOWS%%\write.exe

Note that there is a \ character before the name of the program.

(c) For all references to files within the QUERCUS directory or its QFILES or MFILES subdirectories, use the text:

%%QUERCUS%%

e.g.
%%QUERCUS%%react.exe

refers to a program in the Quercus directory, while

%%QUERCUS%%mfiles\barrels.mtw

refers to a data file in the MFiles sub-directory. Note: there is not a \ character after the "%%QUERCUS%%" text.

1.5 Bugs

Problems, difficulties and bugs in Install should be reported to us at:

cal@stams.strath.ac.uk

with Subject: CAL INSTALL ERROR.

1.6 History

Part 2:

Teaching Guide for Version 3.1

Using Quercus

Once installation is complete, double click on any of the Quercus icons to start a WorkBook/EexerciseBook. We advise first-time users to start with WorkBook1 "Classifying Variables" as this contains simple instructions how to use the Quercus software. For the workBooks we have used a "tabbed notebook" model. Each workBook has three sections. You can select which section you wish to read by clicking on one of the 'tabs' at the side of the contents screen which appears after the list of objectives. Below is a typical Quercus screen. We have labelled some of its features. Each section has approximately 6 pages, the page forward and page back buttons appear in the bottom right hand corner. The End button will close the section from any page and return you to the contents page.

A typical WorkBook Page (WorkBook 5) showing most of the features of the Quercus software: Menus linking to other applications, navigation buttons etc.

The other buttons are:

Along the top of the Quercus Window, as well as the File menu there is a Minitab menu and a Notebook menu to open Minitab and Write.exe respectively. One important feature not illustrated in this example is the 'hotword' feature. Text which appears in red indicates that if that word or phrase is clicked on more information will appear in a hotword box. Typically hotwords are used to provide the user with the option of obtaining supplementary or more detailed information.

WorkBook 1: Learning About Variables

Sections:

1.	What are .i.variables;? Qualitative vs Quantitative.
2.	More about quantitative data. Play the .i.reaction time game;.
3.	Writing your .i.report;. You will need a formatted disk!

Although the overt aim of this first WorkBook is to allow the students to learn how variables are classified, it is in working through this WorkBook that they learn how to use the Quercus courseware. The use of hotwords and various interactions (text, click-touch, click and drag), are demonstrated. In Section 3 the student is asked to summarise this .i.classification; (fig. 1), this together with examples they supply in response the questions on subsequent pages will be written to a file on their floppy disk.

Figure 1: Section 3, Page 2. Student summarises variable classification as a 'tree'. Feedback text will appear in response to incorrect text.

Exercise Book 1: Classifying Variables

At the end of the multiple choice test the students are given a brief summary of their results. They have the option of reading a more detailed report and of having this report written to a text file on their floppy disk. We advise that students should print out this report and keep it with their notes.

**********************************************************************************
 		Feedback from Multiple Choice Test.
 
		exbk1.app
 
		MMCC		25/10/95
**********************************************************************************
Question 1
Incorrect.  In Q1 you were asked to classify a variable describing the reaction to an ointment applied to 
the skin.  Because the reactions were ranked this is an ordinal variable.
 
Question 2
Correct, number of peas in a pod is a discrete variable.
Correct, nMoles mg-1 of ADP in a sample of plant tissue is a continuous variable. 
Correct, the % area of a bacterial plaque in a petri dish is a continuous variable.
Correct, Units of alcohol per week drunk by students is a discrete variable.
 
Question 3
Incorrect, the top-right diagram is the odd one out because the variable on the horizontal axis is 
qualitative.  The variables on all the other axes are quantitative.
 
Question 4
Incorrect a nominal variable can have more than two values, e.g. in an exit poll during an election, the 
variable will be 'candidate voted for'.
Correct, that the data are ranked, or that an order in implied is a characteristic of ordinal data.
Correct, discrete variables may have fractional values.
Correct, an observation of a continuous variable may be expressed to the nearest whole number.
 
Question 5
Incorrect, Graded response to a drug treatment is a qualitative variable, the others are quantitative.
 
Question 6
Correct, a data set is a collection of observations of a variable.
Incorrect, an object which has only one state or value is a constant.
Incorrect, qualitative variables consist of a number of categories which describe the observations, for 
this reason qualitative variables are sometimes called categorical variables.
Incorrect, whether a variable is qualitative or quantitative is less important whether it is appropriate and 
informative in the context of your study or experiment.
  
Consider your test results.  At the beginning of WorkBook1, 'Learning about Variables', it was stated 
that at on completion of the WorkBook you should be able to demonstrate how variables are classified. 
Are you satisfied that you have achieved this objective?  If not we suggest that you; (i) revise 
WorkBook1, (ii) read the appropriate chapters in the recommended text book, (iii) discuss your 
difficulties with fellow students and/or your tutor.

WorkBook 2a: Introduction to Minitab -
Part I: Minitab Basics

Much of the material in this WorkBook was designed for students who have had little or no experience with computers. Tasks such as arranging windows, use of menus, scroll bars and arrow keys are described in detail often accompanied by animations or diagrams. Students already familiar with these tasks will get through this WorkBook quickly.

Figure 2a. Section 2. Page 4. The structure of the Minitab worksheet and the process of data entry are described in detail for novice users.

Exercise Book 2a: Introduction to Minitab I

In section 1 the students; get to practice typing data into the worksheet, are introduced to the concept of subcommands and are encouraged to explore the Minitab help system.In section 2 they are tested on their understanding of how Minitab commands work (see below).

Figure 2(I). Multiple choice True/False question. The students click on either the T of F boxes for each part of the question. Their choice is indicated in the third column and in the answer strip below.

WorkBook 2b: Introduction to Minitab -
Part II: Saving & Retrieving Files

Although the concept of file pathnames is explored briefly at the beginning of section 1 it would be useful if this knowledge was a prerequisite to doing this workbook. The WorkBook deals with all the basic .i.file handing; skills that the students will need; reading and writing ASCII files, saving and retrieving Minitab worksheets, cutting and pasting from Minitab to a text file. We recommend that the students are encouraged to make notes of the relevant commands. The content of the WorkBook is greater than that of WorkBook 1 or 2a. Our experience is that the students require more time to complete it.

Figure. 2b. For all commands in this WorkBook both the 'command line' and menu methods are shown. The students choose which one to use.

Exercise Book 2b: Introduction to Minitab II

More practice saving and retrieving worksheets and reading and writing data files.

WorkBook 3: Visualising Data

In this WorkBook the emphasis is on why and how data is presented in a visual form. The relationship between presentation of data in tabular form and creating a plot is examined in detail using the stem-and-leaf plot as an example. This we hope, will help students understand how to interpret graphical presentations of data. They get a chance to practice this in section 3 where the process of writing a report is outlined. The students are questioned about the aim of the experiment used as an example in the WorkBook and how they interpreted the data (Fig. 4). Their responses are formatted and written as a specimen report to a file on their floppy disk.

Figure 3. Section 3. The students are questioned about how they explored the data

Exercise Book 3: Visualising Data

In the WorkBook the students created histograms, boxplots etc. of the Steroid data. In order to use the graphical output from MINITAB in their reports however they nedd to customise them, i.e. add titles, label axes etc. The sub-commands for customising histograms are demonstrated in section 1 and the students are encouraged to practice using these commands. The multiple choice test in section 2 concentrates on the students understanding and ability to interpret graphs and plots.

WorkBook 4: Measuring Location

In this WorkBook we introduce the idea that data sets have characteristics; location, spread and skewness (Fig. 5), and that 'statistics' quantify these characteristics. The students work on two sets of data: 1. The 'steroid' data set in the Minitab worksheet steroid.mtw. 2. The reaction time data which they generate when they play the .i.reaction time game; (access through Reaction menu). This data is stored as text on their floppy disks. Knowing how to retrieve Minitab worksheets and 'read' ASCII files into Minitab are therefore essential pre-requisites for this WorkBook.

Figure 4. Section 1. Page 7. There is a short multiple choice test to reinforce students recognition of skewness.

Exercise Book 4: Measuring Location

In section 1 the students are asked to load the data set Vanilla.mtw into Minitab and explore the distributions of data in this worksheet. The data are measurements of the amount of vanillic acid in three different types of vanilla pods. The multiple choice questions in section 2 test the students ability to select appropriate measures of location for different distributions.

WorkBook 5: Measuring Variability

The emphasis in this WorkBook is on how statistics (standard deviation, semi-interquartile range), measure the variability within a data set (Fig. 6). This is to help the students appreciate how skewness or the presence of outliers in a data set can affect the value of these statistics. Students meeting these concepts for the first time are liable to find this WorkBook (particularly Section 2), quite difficult.

Figure 5. Section 2. Page 6. In this test the students are asked to demonstrate their understanding of the standard deviation either by repositioning the data points so that the standard deviation is changed while the mean remains constant or vice versa.

Exercise Book 5: Measuring Variability

In section 1 the students retrieve the vanillic acid data and explore the distibutions visually before choosing appropriate descriptive statistics and comparing variability. The interpretation of the results is also discussed. The multiple choice questions test the students ability to select appropriate measurs of variability based on the shapes of distributions.

WorkBook 6: Sampling Theory

Between WorkBooks 5 and 6 the pace and level of difficulty of the course moves up a gear. Again the emphasis is on getting the students to understand the concepts underlying the statistical methods. In this WorkBook Minitab is used not only to explore data, but to simulate data. This data is then used to showing that the Standard Error is a good approximation to the sd of the sampling distribution (Fig. 7).

Figure 6. In section 1 the students get to select their own sample and compare the statistics with those from the “population”..

Exercise Book 6: Sampling Variability

In section 1 the students get the opportunity to explore the exponential as well as the normal distribution. Again they are asked to use Minitab to simulate drawing random samples from pre-defined populations.

WorkBook 7: Confidence Intervals

When testing this WorkBook the students reported that they knew what a confidence interval was and how to use Minitab to calculate it, however they were not sure about the interpretation of confidence intervals (Fig. 8), even after doing the practice exercises in section 3. The work in ExerciseBook 7 should help them. Note that the t-distribution is mentioned in Section 1 but not explained in detail. You may wish to add information about this to the Text file or in a 'handout'.

Figure 7. Section 2. Page 3. This question is based on a common misconception students have about confidence intervals.

Exercise Book 7: Confidence Intervals

WorkBook 8: Significance Testing

How to formulate and test hypotheses is outlined in the first section (Fig. 9), and the students are taken set by step through these processes in section 2. If the students have understood and retained the information about sampling and t-distributions which was presented in WorkBooks 6 and 7 they should be able to understand the material in these sections. The third section deals with Type 1 and Type 2 errors and 'power'.

Figure 8. The Minitab output will give the student a value of t and p but s/he must determine whether or not to accept or reject the null hypothesis.

Exercise Book 8: Significance Tests

WorkBook 9: Regression

In section 1 the idea of exploring the relationship between two variables, first visually, then statistically is developed. In an interactive simulation the students can move a line around the screen to find the one that best models data points on a (Fig. 10). The concepts of fit and residuals are introduced. Section 2 concentrates on defining and calculations of the coefficients of a simple regression model. In Section 3 the idea of confidence intervals for predicted y-values is explored.

Figure 9. A simulation in which students can determine the best line to describe the data.

Exercise Book 9: Regression

WorkBook 10: Correlation

In the first section the we explore the statistical meaning of 'correlation'. The concept of the correlation coefficient 'r' and how it's absolute value and sign indicate the nature and strength of the relationship between two sets of data is explored. In section 2 we simulate an experiment based on the Muller-Lyer illusion test to allow the students to create their own data sets (Fig. 10). Being the subject of the experiment we hopefully give them some useful insights when they are analysing the data in Section 3.

Figure 10. The students are asked to explore the relationships between the actual and their estimates of the lengths of the adjustable arrow in the Muller-Lyer illusion test.

Exercise Book 10: Correlation

WorkBook 11: Residuals and Transformations

In section 1 the concept of using plots of standardised residuals to explore the validity of linear models is introduced. The students are shown a number of different ways of plotting residuals and their interpretation is discussed. In section two we deal with modelling simple, non-linear relationships. How to select an appropriate transformation and how to transform the x-variable are discussed. In the third section the students are given 4 sets of data relating to a plant competition experiment, and asked to consider 4 pair-wise relationships between them. In two cases the relationships are reasonable linear, in the third it is curved. The shape of the fourth relationship is debatable, the students will need to make a choice and consider the biological interpretation of whichever model they choose.

Figure 11. Students choose a transformation for x by clicking on one of the button on the right-hand side. The points then move to new positions.

WorkBook 12: ANOVA

In section 1 the students are introduced to the idea of partitioning variability according to identifiable sources and using the F-test to determine if the variability between treatments is significant. We do not illustrate how to do the calculations for an ANOVA (although the MINITAB commands are shown in sections 2 & 3), but concentrate on interpreting the results. The linear model for ANOVA is discussed in relation to experimental design. In Section 2 the students analyse data from a completely randomised design. In section 3 the concept of blocking is introduced and an example of a a randomised block design explored.

Figure 12. Illustrating the MINITAB command for analysisng data from an RBD.

Index