Centre for

Statistical Education

Database
Interrogation Tool
Beta Version

User Guide

Database Interrogation Tool

Introduction
The Database Interrogation tool or 'Datatool
has been designed to allow quick and easy
access to the large amount of data collected
by the AtSchool projects such as
CensusAtSchool and Experiments AtSchool
The tool works through a web browser such as
Internet Explorer and requires a live internet
connection so that data can be requested and
sent from the AtSchool databases.

The browser software that you use will also need to have Adobe Flash (v8 or later) installed.
Updates of Flash can be obtained for free from www.adobe.com.
Selecting a dataset
The online tool can currently connect to one dataset - CensusAtSchool. Click on the
appropriate logo to show the selection of results for this AtSchool project.
Click on the name of a particular questionnaire
or survey you wish to investigate:

Your selection will be marked with a tick:

It is not possible to select multiple

questionnaires at the same time.

Once a questionnaire has been selected you will be to move on to the next screen.
To navigate between screens select either the Next button (top right) or the Back button
(top left) :

On certain screens, the Next button will only become visible once you have selected sufficient
options.

Database Interrogation Tool

If you want to find out more about a
questionnaire or experiment, click on the
circled question mark:

A new browser window will appear which has

details related to the selection
For CensusAtSchool you will then be able to
access the Adobe PDF and/or MS Word
documents related to the relevant
questionnaire.

Selecting a variables

The third screen will show a list of variable

names held within the database you have
selected.
The list of variable names is colour-coded.
Yellow items, or tabs, represent continuous
variables and discrete variables names appear
as red tabs:

The coloured tabs can be dragged around the screen and

dropped on to the list of variables to be investigated (collected
variables). You can investigate a maximum of six collected
variables in one go.
You may mix and match continuous and discrete variables in
your selection. Please note the datatool may re-order the list of
variables you select to remove blank cells.

Once you have finished making your

selection move onto the next screen.

Database Interrogation Tool

Selecting a random sample
You can then select the type of random
sample you wish to collect. This screen allows
you to change the size of the sample and to
collect data for a particular category/group.
The size of the sample, that can be collected,
can be varied from 50 to 250. The default
sample size is 100.
Please note that larger samples may take
longer to collect and analyse depending on the
speed of your internet connection and
processor power of your PC.

Selecting one or more samples

In an investigation, you might want to collect
and compare data for two groups for
example data from just Girls with the
equivalent from Boys. On the left hand side
of the screen, select the Two samples
option.
You will now be able collect two samples; in the interrogation tool these are referred to as
sample A and B. Once you have chosen the features you want for sample A, clicking on the
Next button allows you to choose features for sample B.
Selecting a sub-sample
As a possible investigation, you might decide
to collect data for a specific year group and
gender.
To do this, select the Part of database tab.

If you have selected the Part of database

option, when you move on to the next screen,
a new list of the variable names from the
database will appear. The new tabs work as a
clickable buttons rather than as dragable
items.
A tab with a star on it indicates that it is one
of your collected variables. You should only
use the starred tabs when you want the
collected variables to satisfy particular
conditions, for example when you only to
collect heights in the range of 100 to 200cm.

Database Interrogation Tool

Suppose you wish to obtain a random of sample of heights and foot length data for boys aged
14. First click on the Gender tab. This will show a list of options on the left hand side of
screen.
Select the Male option. As you do so, the Gender tab will be ticked to show that you are using
it as a filter.

Click on the Age tab; the maximum and minimum values will appear in the option list. Alter
the maximum and minimum values to be 14 and then click on the Save this option button.

The interrogation tool will now select a random sample of boys aged 14 and make these
available for processing. To remove a filter, click on the Cancel this option button.
Collecting the data
Once you have finished making your selection of random sample move onto the next screen.
The datatool will communicate with the online database and collect your sample.

Once the requested data has been successfully returned the tool will automatically move on to
the next screen.

Database Interrogation Tool

Displaying the data
After the data has successfully returned from
the online database, it can be displayed in
tables and charts.
The data can also be copied to your
computer's clipboard to be exported into a
document/spreadsheet or presentation.
The raw data is not automatically shown.
Instead a choice of charts is given which can
be used to display the data.
To view the raw data, select the magnifying
glass icon at the top of the Variables list.
This will open up a screen of Tabulated data with information on the Random sample, a
Frequency chart and statistical measurements such as Mean, Range and Standard deviation.

By default, information for the first item in the Variables list is shown. To select another item,
click on the name of the variable in the list on the right hand side of the screen.
To display the data graphically, select the Back button to return to the choice of charts.

Each tab in the list of Variables can be dragged and dropped on to the chart icons. Beneath
each chart is a coloured box indicating the type of variable that it will accept.

Database Interrogation Tool

Histograms, Box-whisker, Dot
plots, 3D colour plots, and
Starplots can only be plotted
for continuous variables:

Pie charts and Bar charts can

only be plotted for discrete
variables:

Scatter graphs and

Tabulated data can be
plotted using a mixture of
variables:

Once an appropriate type of variable has been dropped onto a chart, an arrow button will
appear:

Click on this arrow to view the chosen chart:

If you have collected two samples, several of the charts and tables
allow both sets of results to be plotted simultaneously.
Where this option is not available you can swap between samples
by selecting one of the tabs shown opposite.

will be
If you have used a sample filter, an information button
added underneath the chart. Clicking on this button will reveal a
pop-up window showing the particular sub-samples you used.

Database Interrogation Tool

Charting Options

All charts have some additional options. Click on the Show options button on the top menu
bar.
Bar charts and Pie charts
These can be arranged in descending order (by percentage) or arranged in
alphabetical/default order. Select or deselect the Arrange in order button and then OK.

As with Bar charts and Pie charts, options for the layout of the other types of chart can
be accessed using Show Options button on the top menu bar.
For Histograms, you can change the range of values plotted and the number of bins these alterations will only take place if you select User defined scale and click on OK.

Database Interrogation Tool

When you plot a histogram for two samples, each will be drawn back to back. Altering the
option values will alter both charts simultaneously.

Box and whisker plots (Boxplots)

You can choose to show or ignore outliers. If the Calculate outliers option is not selected,
the whiskers will extend from the ends of the box to the maximum and minimum values.

Database Interrogation Tool

Scatter graphs
Scatter graphs can be plotted with continuous or discrete values along the x-axis.
(The variable selected for the y-axis must be continuous.)
You can choose to alter the range covered by either axis, the transparency of the plots.
When using a high transparency most data points will appear faint except in cases where
several points, with the same co-ordinates, have been plotted on top of each other.
When both x- and y-axis use continuous variables, both a line of best fit and the related
equation (y= a + bx) can be included on the chart:

When discrete data is used on the x-axis, boxplots can be overlaid on each series of points:

Database Interrogation Tool

Tabulated data
In the example below, a table has been created using Live In with Height. For the discrete
variable, Live In, four columns are created to match the four possible values (City, Town,
Village, Other). These are fixed.
However for the continuous variable, Height, the number of rows can be varied using
Show options on the top menu bar. The user can alter the number of bins (rows) and
range covered.

Dot plots
Dot plots can only be plotted with continuous data. The space beneath the chart can be
used to show a boxplot.

Database Interrogation Tool

Starplots
Unlike the other chart types you cannot drag and drop the names variables on to the
Starplot graphic.

The arrow button for this particular

graphic will automatically appear when
three or more continuous variables have
been collected.
Clicking on the chart button will load
another option screen

Each Starplot consists of equi-angular spokes or 'arms', with each spoke representing one of
the continuous variables. The length of each arm is based on how each individual
measurement (variable) compares with the maximum value for the sample - i.e. each arm
represents a ratio within the range 0-1.

The default Starplot shows the maximum length possible for each measurement by adding
thin lines in grey beneath each arm. The transparency of these arms can be altered:

Database Interrogation Tool

3D colour plot
Unlike the other chart types you cannot drag and drop the names variables on to the
3D colour plot graphic.

Clicking on the chart button will load another option screen:

Drag and drop the names of continuous variables on to the x-, y- and z-axis. When all three
axis are filled, the chart can be plotted.
The following example uses data from CensusAtSchool Phase 4. Pupils were asked how
much they like PE, Maths and English using a continuous scale in the range 0-5 where 0 =
'Do not like' and 5 = 'Like'.
For the random sample, pupils' ratings for these three subjects are used to create (x,y,z)
co-ordinates for a 3D plot.
Furthermore the (x,y,z) values are used to create colours for each point. A high x-value
adds red to the individual's point. A high y-value adds green and a high z-value adds blue.
Therefore a student at coordinates (5.0, 0.3, 4.9) will be coloured purple.
Using this RGB (red-green-blue) mixing, a student with (5,5,5) will have a point coloured
white and a student with (0,0,0) will have no colour i.e. black.

Database Interrogation Tool

Interactive Charts
As the mouse moves over parts of particular charts, values can be displayed next to the cursor
by clicking on the point and holding down the mouse button.
For example values of outliers are identified in this way on Box and whisker plots:

The 3D colour plot can be rotated around. Click on the chart, hold down and move the mouse the chart will rotate according to how the mouse is moved (up/down and left/right):

Database Interrogation Tool

Storing Charts
Up to four charts can be stored in the Datatool's memory as you use the program. Move the
mouse over the Save charts icon in the top right hand of the screen. A folder with four spaces
will appear. Click on the particular space inside the folder you wish to use.

Click on View saved charts icon and the folder will automatically open up to reveal the charts
you have stored:

To review a stored chart, click on the thumbnail and hold down the mouse button. Release the
mouse button to minimize the chart. As each space is used a tick will appear. Used spaces can
be overwritten with new charts if you wish.
The copies of charts you place inside the folder are no longer active they are purely low quality 'screen grabs'. It is hoped that a future
version of the tool will allow you to save these charts to a local
folder/hard drive.
To create a high quality screen grab, which you can paste into an
external document, click on Save charts and the select the Camera
icon. This will clear the screen of buttons to leave only the chart. You
can then use screen capture software or Alt+PrtScr. Once you have
captured the chart, click anywhere on the Datatool screen to reveal the
buttons/menus.