Poster Presentations
Creating Posters
The following Web sites contain information about successfully creating posters for presentation:
- Creating Posters using Power Point
- Creating a Poster Using MS PowerPoint
- Ten Steps to Better PowerPoint Posters (.pdf file) from Media Resources at the College of London
- Creating Posters in Power point (.pdf file); created by Richard Scott
- Resources and Links to Information about Creating Posters using PowerPoint
Student Committee Poster Presentation Quality Assessment Results
Background
At the December 2002 meeting of the Executive Council, it was proposed that the Student Committee generate a plan to assess the quality of poster presentations at the May 2003 Conference. In response, the Student Committee generated a proposal, which was approved by the Council prior to the May 2003 Conference. The results of the quality assessment are reported below, followed by recommendations for improving the quality of posters at future conferences.
General Procedures
Of the 701 posters at the conference, 89 (12.7%) were selected for evaluation by Student Committee volunteers. Volunteers rated the posters using a 28-item checklist, which included a subjective measure of quality (Appendix A). To ensure that the evaluated posters were representative of all posters at the conference, posters were selected at random from each of the 12 presentation categories, including at least 10% of the posters in each category. For example, there were 151 posters in the Autism category, therefore 17 (11.3%) were selected. In contrast, there were only 31 posters in the Teaching Behavior Analysis category, therefore only 5 (16.1%) were selected.
Of the 89 posters selected, only 61 posters were evaluated at the conference. The fewer number of posters evaluated was due to posters not being displayed, volunteer attrition, and logistical problems. Figure 1 displays the number of posters evaluated in all conference categories. Although a fewer number of posters were evaluated than planned, at least one poster was evaluated in each category. Further, the number of posters evaluated was generally proportional to the total number of posters in each category.
Figure 1. Number of posters evaluated in all convention categories.
Interobserver Agreement
To establish believability of the data, two evaluators completed the rating checklist for 14 (23%) of the posters. Percentage of interobserver agreement was calculated by dividing agreements for all 28 items by agreements plus disagreements and multiplying by 100. Average interobserver agreement was 72.7% (R= 53.6% - 89.3%).
Results
Figure 2 displays the average percentage of “YES” responses for checklist items. The bar labeled “Overall” represents the percentage of “YES” responses for all checklist items, while the remaining bars represent the percentage of “YES” responses for each category. On average, posters met 75.5% (R=29%-96%) of the quality criteria identified in the checklist. The categories with the highest percentages of “YES” responses were Title (89.2%) and Author’s Name (95.2%). The categories with the lowest percentages of “YES” responses were Text Body (59.5%) and Captions (59.7%). The two remaining categories, Figures and Layout, had 75.9% and 80.5% “YES” responses, respectively.
Figure 2. Average percentage of “Yes” responses
for rating checklist items, overall and by category.
Table 1 shows the average percentage of “YES” responses for each item on the checklist. Reflecting the data in Figure 2, the Title and Author’s Name categories met the highest percentages of rating criteria. For example, 97.1% of posters included the first and last names of authors in the title, and 95.5% of poster titles were visible from six feet. The Text Body, Captions, and Figure categories had considerably lower ratings. Over 50% of the posters used more than six lines of text in a paragraph and had more than 250 words of text in the entire poster. Further, over 45% of posters did not present captions for all figures and over 50% did not have figure labels, text and scales that were visible from at least 6 feet. Together, these data suggest that many posters contained too much text and had figures, including graphs, that were difficult to read.
Category |
Item |
% Yes |
Title |
Title is legible from 6 feet | 95.5% |
| If the title is more than one line, title capitalization is used | 84.1% | |
Author's Name |
Author(s) names are complete | 97.1% |
| Title capitalization is used | 92.7% | |
| Author’s affiliations are listed | 94.9% | |
| Authors are grouped by affiliation | 92.7% | |
Text Body |
Text is legible from 6 feet | 81.9% |
| A Sans Serif font is used | 58.3% | |
| 45 letters per line, maximum | 65.6% | |
| 6 lines in a paragraph, maximum | 49.3% | |
| 250 total words, maximum | 46.4% | |
Captions |
All figures have captions | 53.9% |
| Captions are descriptive | 72.9% | |
Figures |
4 bar graph categories or 4 lines of data, maximum. | 92.6% |
| Data points on line graphs are clearly visible from 6 feet | 78.7% | |
| No lines in the background of graphs | 70.0% | |
| Multiple data lines have different data points or colors | 90.0% | |
| Multiple categories on a bar graph have different fill or colors | 77.1% | |
| If single-subject data are presented, line graphs are shown | 95.9% | |
| Horizontal and vertical axis labels are present on line graphs | 93.6% | |
| Labels are legible from 6 feet | 47.7% | |
| Text and scales are self-explanatory | 84.8% | |
| Text and scales are legible from 6 feet | 39.7% | |
| If more than one color is used, the contrast is sufficient | 73.8% | |
| Grouped data includes measure of variability (e.g., error bars). | 46.7% | |
| No more than 1 table total | 87.6% | |
Layout |
All elements in landscape | 64.8% |
| Poster is ordered in titled sections (conclusions, methods, results, etc.) | 93.2% |
Table 1: Average percentage of “Yes” responses for rating checklist items by item.
Figure 3 displays the average percentage of subjective ratings for all posters. 56.7% of posters received an excellent or good rating. 35% of posters received an average rating, while only 8.4% received a below average or poor rating. The subjective data indicate that, on average, raters perceived the large majority of posters to be average to excellent in quality.
Figure 3. Average percentage of subjective ratings for all posters.
Discussion and Recommendations
Data from the quality assessment suggest that poster presentations at the May 2003 ABA Conference had several strengths. First, the title, authors, and affiliations for most presentations were readable from a distance of six feet. An easy to read title allows the viewer to quickly understand the topic of the presentation, and invites her to inspect the text and figures more closely. Second, letters in the text body for most presentations were visible from six feet. Highly visible letters aid the viewer in understanding the content of the presentation. Third, the large majority of presentations were ordered into labeled sections, which also facilitates readability. Finally, subjective assessments indicated that, despite some flaws, the large majority of posters were average to excellent in quality. The subjective data reflected that most viewers were at least satisfied with the quality of presentations.
The data suggest that poster presentations could be improved in a number of ways. The following are specific areas for improvement:
- Presenters should use a Sans Serif font in the poster, which is more readable than Times New Roman or other font.
- Less text should be printed in the poster. Viewers are unlikely to read long sentences and lengthy paragraphs. Presenters should limit text to less than 45 letters per line and less than 250 words in the entire poster.
- Captions should be presented for all figures. Captions should clearly describe what the figure represents.
- Figures and figure text should be readable from at least six feet. Labels, scales, data points, and data paths in figures should be clearly readable.
- All elements of the presentation should be in landscape format.
- If appropriate, a measure of variability should be presented for grouped data.