Producing top quality tables and figures in APA format requires a bit of thought and effort.  These instructions are designed to make the process a bit easier.

Figures

A well-prepared figure can convey the qualitative aspects of data, such as comparisons, relationships, and structural or pictorial concepts, more efficiently than can text or tables.

Standards for figures. The standards for good figures are simplicity, clarity, and continuity. A good figure
* augments rather than duplicates the text
* conveys only essential facts
* is easy to read, with elements (type, lines, labels, etc.)
* is large enough to be read with ease
* is easy to understand--its purpose is readily apparent
Types of figures. Graphs show relationships--comparisons and distributions--in a set of data. There are four major types of figures, line graphs, bar graphs, scatter graphs, and photographs.
1. Line graphs are used to show the relation between two quantitative variables. They are most often used to show continuous change or when the shape of the curve or curves is important.

2. Bar graphs or histograms are simple, adaptable, and telegraphic. They are used when the independent variable is categorical (e.g., as with different experimental conditions).

3. Scatter graphs consist of single dots plotted on a line graph; the dots are not joined by lines. A cluster of dots along a diagonal indicates a correlation.

4. Photographs have excellent eye appeal. They should be of professional quality and should be prepared with a background that produces the greatest contrast.

Many of the design and execution problems involved in producing figures are reduced with the use of the computer. There is a description of several common softwarre packages that produce figures in the Psych600 folder (Plotting.wp).




Creating Figures with Excel

One of the easiest ways to produce top quality figures is to use the spreadsheet Excel. Although the default settings for charts do not meet APA format standards, just a few modifications of the figure will produce a top figure.

There are two basic types of figures that are used in psychology papers, line graphs and bar charts. It is possible to produce both types using Excel. In fact, most of the procedures for one type can be used with the other type as well.

Line Graphs

One of the most common figures used to present psychological work is the line graph. This type of figure is usually used when the variable on the X-axis is continuous.

1. Getting Started

Let's assume that we want to produce a line graph or bar chart of the interaction between the two variables in your study. The study involved looking a male and female college students who were given three blocks of trials using a new typing system. The researcher was interested in how learning (determined by the number of correct responses) changed over the trails.

The means for each of the six conditions were determined and then entered into an Excel spreadsheet. There are certainly many ways to do this but a convenient format is shown to the right. Note that we want to include labels as well as data.

2. Chart Wizard

The first step in creating any type of figure in Excel is to make use of the Chart Wizard. Highlight all of the data, including the labels. Then engage the Chart Wizard.

Chart Wizard Icon

The initial choice that the Chart Wizard provides is the type of figure wanted. Select line from the Standard Types then click Next once then Next again (we don't need to change the options provided). Note: If the figure does not look like you wish with the line option, you may want to try the XY Scatter option and then choose the Scatter with data points connected by lines (option in the bottom left corner). This option will often produces a better looking plot if you have more than two lines of data. 

The third Chart Wizard screen provides an option to add labels to the X and Y-axes. For the X axis type 'Trial Number' while for the Y axis use 'Mean Number of Correct Responses.' Do not enter a title for the chart, we will do that later.

Click Next and then Finished.

The rough version of our figure is now placed on the spreadsheet. However, there is still much to do.

3. Chart Toolbar

Virtually all of the modifications that we will now do to the figure will be done using one of the options from the Chart Toolbar (View/Toolbar/Chart, then use the pull-down menu to select the chart feature you are interested in). Here is an example of a Chart Toolbar.

In this case this toolbar is associated with the Legend on the figure.

4. Options

The major options for this part of the chart are displayed by clicking the Options icon. (There is a second may to get Chart Toolbar Options, just double click on the area of interest. For example, if we double clicked on the Legend, we would bet the Legend Toolbar Options.)

A) Borders and Lines

The first thing we need to do is to eliminate the extra borders and lines in the figure. Bring up the Plot Area Options and change the settings as follows:

            - Border: none
            - Area: none

Then click OK.

On the Value Axis Major Griding Options, select Custom color then the box with the color white. Again, OK.

These changes should eliminate the borders and griding from the figure. Your figure should now look like this.

B) Where to put the Legend?

In APA format the legend is usually placed within the figure, not to the side as Excel puts it. But before we can move the legend inside the figure, we need to provide more room. There are two parts of the figure which can be sized, the overall picture (lines, text, and legend) and the smaller area that just contains the lines of the graph itself.

Begin by selecting the larger area (click somewhere inside it) and make it both wider and longer (click the small black boxes, hold down the left mouse key, and drag). Now make the smaller central area a bit larger. This should provide space for the legend. Select the legend and then move it to an appropriate place inside the figure. (As you change the size of the figure, keep in mind that the Y axis should be 2/3 the length of the X axis.) You now should have something that looks like this.

Before we leave the Legend, note that we have removed the boarders from the Legend. To do this select the Pattern option and then Borders: none.

C) Font Size for Labels

The font size and type of all of the labels on the figure (Mean Number…, Trial…, 0, 25, …, Male, etc.) can be changed to give the figure a more balanced look. Just click on the label you want to modify, select the appropriate font size and type, and click OK. (A font size of 10 or 12 is usually appropriate. The font style used for this example is Times New Roman.)

D) Line Type and Symbols

The lines themselves can be changed in a number of ways. Double click on one of the line segments or on one of the end points to bring up an Options panel. Select the Patterns tab. You can change the line type (dotted, solid, etc.) or width. The type of symbol used for each line (called a Marker) can also be changed.

Change the lines and markers until you can clearly distinguish between them. Remember that your final product will be in black and white so color can not be used to differentiate the lines.

E) Ticks

APA format suggests that the small lines on the X and Y-axes, called ticks, should be on the inside, rather than the outside, of the lines. Get the Category Axis Options and change the ticks on Y-axis from outside to inside. Change those on the X-axis from outside to none.

F) Error Bars

The final items we need to add to our figure are error bars. Error bars are used to indicate the spread of the data and are used to represent plus/minus 1 standard deviation, plus/minus one standard error, etc. Note: Error bars should not automatically be added to every figure.  Include them ONLY when they are needed.

We need to use the same Options panel as we used to change the line type. Just double click on one of the lines. Select the Y Error Bars tab.

Since we entered the means and not the raw data in our spreadsheet, we can not use Excel to calculate the standard deviations or standard errors. Rather we must use the Fixed value option and enter our standard deviation or standard error which will then be displayed on the figure when we click OK.

However, before you click OK, make sure that the proper Display window is selected. Since we want plus/minus, we select that window.

What if I want error bars of different length for each point?  Select the Custom option and then enter the values you need with a comma between them.  For example, if you have four points and the error bars should be 6, 4, 5, and 4, just enter 6,4,5,4.  If you want both + and - error bars, you will need to enter the values twice, once in each box.

6. The Final Product

The completed figure should look like this.

7. Putting the Figure in Your Paper

Once the figure is done to your satisfaction, just copy and paste it into your Word document. It is at this final stage that the title of the figure can be added at the bottom of the figure. Your final product might look like this.

Figure 1. Mean number of correct responses as a function of trial number of 20 male and 20 female college students.
 

Bar Charts

In presenting the results of an interaction, it is often more appropriate to use a Bar Chart rather than a line graph. (According to the APA Style Manual, line graphs should only be used when the independent variable on the X-axis is a categorical variable.)

1. Getting Started

Let's assume that we want to produce a line graph or bar chart of the interaction between the two variables in your study. The study involved looking a male and female college students who were given three blocks of trials using a new typing system. The researcher was interested in how learning (determined by the number of correct responses) changed over the trails.

The means for each of the six conditions were determined and then entered into an Excel spreadsheet. There are certainly many ways to do this but a convenient format is shown to the right. Note that we want to include labels as well as data.

2. Chart Wizard

The first step in creating any type of figure in Excel is to make use of the Chart Wizard. Highlight all of the data, including the subject headings. Then engage the Chart Wizard.

Chart Wizard Icon

The initial choice that the Chart Wizard provides is the type of figure wanted. Select column from the Standard Types then click Next once then Next again (we don't need to change the options provided).

The third Chart Wizard screen provides an option to add labels to the X and Y-axes. For the X axis type 'Trial Number' while for the Y axis use 'Mean Number of Correct Responses.' Do not enter a title for the chart, we will do that later.

Click Next and then Finished.   The rough version of our figure is now placed on the spreadsheet. However, there is still much to do.

3. Chart Toolbar

Virtually all of the modifications that we will now do to the figure will be done using one of the options from the Chart Toolbar (View/Toolbar/Chart, then use the pull-down menu to select the chart feature you are interested in). Here is an example of a Chart Toolbar.

4. Options

In this case this toolbar is associated with the Legend on the figure.

A) Borders and Lines

The first thing we need to do is to eliminate the extra borders and lines in the figure. Bring up the Plot Area Options and change the settings as follows:

        - Border: none
        - Area: none

Then click OK

On the Value Axis Major Griding Options, select Custom color then the box with the color white. Again, OK.

These changes should eliminate the borders and griding from the figure.

B) Where to put the Legend?

In APA format the legend is usually placed within the figure, not to the side as Excel puts it. But before we can move the legend inside the figure, we need to provide more room. There are two parts of the figure which can be sized, the overall picture (lines, text, and legend) and the smaller area that just contains the lines of the graph itself.

Begin by selecting the larger area (click somewhere inside it) and make it both wider and longer (click the small black boxes, hold down the left mouse key, and drag). Now make the smaller central area a bit larger. This should provide space for the legend. (As you change the size of the figure, keep in mind that the Y axis should be 2/3 the length of the X axis.) Select the legend and then move it to an appropriate place inside the figure.

Before we leave the Legend, note that we have removed the boarders from the Legend. To do this select the Pattern option and then Borders: none.

C) Font Size for Labels

The font size and type of all of the labels on the figure (Mean Number…, Trial …, 0, 25, …, Male, etc.) can be change to give the figure a more balanced look. Just click on the label you want to modify, select the appropriate font size and type, and click OK. (A font size of 10 or 12 is usually appropriate. The font style used for this example is Times New Roman.)

D) Ticks

APA format suggests that the small lines on the X- and Y-axes, called ticks, should be on the inside, rather than the outside, of the lines. Get the Category Axis Options and change the ticks on Y-axis from outside to inside. Change those on the X-axis from outside to none.

At this point your figure should look like this.

E) Error Bars

The final items we need to add to our figure are error bars. Error bars are used to indicate the spread of the data and are used to represent plus/minus 1 standard deviation, plus/minus one standard error, etc.

We need to use the same Options panel as we used to change the line type. Just double click on one of the bars. Select the Y Error Bars tab.

Since we entered the means and not the raw data in our spreadsheet, we can not use Excel to calculate the standard deviations or standard errors. Rather we must use the Fixed value option and enter our standard deviation or standard error which will then be displayed on the figure when we click OK.

However, before you click OK, make sure that the proper Display window is selected. Since we want plus/minus, we select that window.

What if I want error bars of different length for each point?  Select the Custom option and then enter the values you need with a comma between them.  For example, if you have four points and the error bars should be 6, 4, 5, and 4, just enter 6,4,5,4.  If you want both + and - error bars, you will need to enter the values twice, once in each box.  However, if you have a figure with two independent variables, you will need to make sure that you have selected a subset of the bars (i.e., those for one of the variables) and then enter the error terms for that set only, repeating the process for the second independent variable. 

F) Bar Patterns

We can use the same panel as we used to display error bars to change the pattern of the bars so they are distinguishable. (Remember that your final product will be in black and white, not color.)

Select the Patterns option the Format Data Series panel. Chose the Automatic option from the left side of the panel to provide the bars with a border.

In the Area section of this panel (right side), select Automatic then Fill Effect. There are many ways you can 'fill' your bars but for a scientific paper the best option is Pattern. If you have two bars in each group, selecting vertical lines for one group and horizontal for the other is often the best option.

5. The Final Product

Your final product should look something like this.

6. Putting the Figure in Your Paper

Once the figure is done to your satisfaction, just copy and paste it into your Word document. It is at this final stage that the title of the figure can be added at the bottom of the figure. Your final product might look like this.

Figure 1. Mean number of correct responses as a function of trial for 20 male and 20 female college students.


Tables

An informative table supplements--it does not duplicate--the text. In the text, refer to every table and its data. In the text tell the reader what to look for in the table, discussing the table's highlights. If you discuss every item on the table in text, the table is unnecessary.

Tables must be intelligible without reference to the text. Explain all abbreviations (except such standard statistical abbreviations as M, SD, and df). In the text, refer to tables by their numbers:

...  as shown in Table 8, the responses ...

... children with pretraining (see Table 8) ...

Do not write "the table above/below" or "the table on page 32."

When preparing tables, all number should be reported using the same number of decimal places. For example, instead of 3.4, 2.56, and 1, you should report 3.40, 2.56, 1.00.

Table numbers. Number all tables in the text with Arabic numerals in the order in which the tables are first mentioned in the paper. Identify tables of the appendix with capital letters and Arabic numerals (e.g., Table A-1 is the first table of Appendix A, Table C-2 is the second table of Appendix C).

Table titles. Give every table a brief but clear and explanatory title. The title should be left justified and either underlined or in italics.

Examples. Samples of well prepared tables are provided in the Appendix. Examine them carefully. Pay special attention to Table 3, an example of an ANOVA table. Note that the table includes columns for df and F but none for SS or MS. MS error is given in parentheses. See your comp advisor for additional help in preparing correct tables.



Placement of Figures and Tables

Until recenlty figures and tables were placed on separate pages in a manuscript. The new convention is that figures and tables may be incorporated directly into the body of a paper, if that is convenient. A figure or a table should appear after the first reference to it. That is, a statement such as, "see Figure 1" should appear before the figure is placed in the paper. If figures or tables are on separate pages, number these pages just as you do all pages in the paper.



Sample Figures

Both of these figures were done using Excel. However, in order to get figures that meet APA style requiresments, as these figures do, it is necessary to move beyond the standard default options.
 
 


Figure 1. Mean number of correct responses as a function of trial number for male and female rats. (n = 25)

Figure 1. Mean number of correct responses as a function of trial number for male and female rats. (n = 25)











Sample Tables

The following are examples of tables that meet APA style.
 

Table 1
Interrelations Between Subscales for Students and Older Adults



Subscale                                                               2                     3                     4

University students (n = 200)

1. Tranquility                                                      .93                  -.09                    .73

2. Goodwill                                                         __                  -.34                    .62

3. Happiness                                                                             __                      .14

4. Elation                                                                                                             __


Older adults (n = 189)

1. Tranquility                                                       .42                 -.07                    .52

2. Goodwill                                                         __                  .43                     .62

3. Happiness                                                                             __                      .47

4. Elation                                                                                                             __



Note: r > .52, p < .05.
 

Table 2
Mean Numbers of Correct Responses by Children With and Without Pretraining


Grade


Group                                                         n               3             4                5

Verbal Tests
Girls

    With                                                     18             280           297            301

    Without                                                19             240           251            260

Boys

    With                                                     19             281          290            306

    Without                                                20             232          264             221


Mathematical tests
Girls

    With                                                     20            201           214            221

    Without                                                17            189           194            216

Boys

    With                                                     19            210            236           239

    Without                                                18             199           210           213



Note Maximum score = 320
 

Table 3
Analysis of Variance for Classical Conditioning


F


Source                                                 df                           Finger CR                     Irrrelevant CR


Between subjects

Anxiety (A)                                            2                           0.76                                 0.26

Shock (S)                                              1                             0.01                                 0.81

A x S                                                     2                             0.18                                 0.50

Error                                                    30                          (16.48)                              (15.73)


Within Subjects

Blocks (B)                                             4                             3.27**                             4.66**

B x A                                                     8                             0.93                                 0.45

B x S                                                     4                             2.64*                               3.50**

B x A x S                                               8                             0.58                                 0.21

Error                                                   120                           (1.31)                               (2.46)



Note: Values enclosed in parentheses represent mean square errors. CR = conditioned response; S = subjects.
* p < .05. ** p < .01.

Table adapted from the Publicaiton Manual of the American Psychological Association, 4th edition (1994). Washington, D.C.: The American Psychological Association.


Contents  |  Psychology   |  Allegheny
5/00