Abstract
The application of established principles from multimedia learning, education, and cognitive psychology can help maintain audience attention during PowerPoint presentations. Slide design should adhere to key cognitive principles, including dual-channel theory, cognitive load theory, and the principles of redundancy, coherence, signalling, and segmenting. Slides generated by AI tools are often overly complex, with redundant text, intricate visuals, and multiple arrows depicting relationships; such slides also cannot be animated. Practical workarounds for these limitations are discussed. Default templates and Design Suggestions frequently use small font sizes that require correction. PowerPoint features such as animation, Draw and Highlight can be used to emphasise key information, while SmartArt facilitates clear representation of relationships between elements. Features such as slide transitions and WordArt are best avoided. Tools such as Pointer, Zoom, and Presenter View help guide the presentation and focus attention. For online presentations, larger fonts and higher contrast are recommended, along with the inclusion of interactive elements such as polls. AI add-ons integrated within PowerPoint offer useful functionality but should be used with care. Narration remains central to effective presentations; slides should support and complement it. The Rehearse with Coach feature may help improve delivery. Supplementary files provide detailed examples of the use of SmartArt and shapes in psychiatric presentations.
Keywords: PowerPoint, presentation skills, slide design
INTRODUCTION
Despite its limitations,[1] PowerPoint offers features that help presenters deliberately direct audience attention to the element most relevant at any given moment. While the option to animate bullet points sequentially has been available since the early versions of the software, features such as SmartArt and Zoom were introduced over the past two decades; however, many presenters make limited use of them. With the increasing adoption of LED displays and online presentations, most presenters now prefer the 16:9 aspect ratio, but often without careful consideration of how to utilize the additional screen space. Furthermore, a recent trend is the use of slides generated entirely by artificial intelligence (AI) tools. These slides are often single images, within which animation is not feasible, and include multiple decorative elements that may hinder learning. An example is provided in Figure 1. A parallel development has been the heavy consumption of short-form video content by younger populations such as students, which has been associated with reduced sustained attention and attentional control.[2,3] Moreover, during online presentations, the speaker must compete with multiple tabs, notifications, and other distractions for the viewer’s attention.
Figure 1.
Example of a complex AI-generated slide
In this context, this article draws on empirical research from multimedia learning, education, and cognitive psychology to propose practical guidelines for slide design. These guidelines aim to help presenters direct audience attention and thereby improve comprehension, retention, and application of the presented material.
RELEVANT COGNITIVE PRINCIPLES
Dual-channel theory: Visual and verbal information are processed through separate channels, each with limited capacity[4]; when one channel is overloaded (e.g., the verbal channel with dense text), working memory capacity is exceeded, causing disengagement.[1]
Cognitive load theory: Slides should be designed to minimize unnecessary processing demands so that limited working memory resources can be devoted to understanding and integration.[5]
Redundancy principle: Presenting identical verbal information as both narration and on-screen text may impair learning by overloading the verbal channel.[6]
Coherence principle: Slides are more effective when extraneous information (e.g., irrelevant images, decorative elements, or non-essential text) is minimized.[7]
Signaling principle: Learning improves when cues such as highlights, arrows, or emphasis are used to direct attention to essential information and clarify the underlying structure.[6]
Segmenting principle: Learners benefit when complex information is presented in smaller, sequential segments rather than as a single dense unit.[6] This allows sufficient time to integrate related elements instead of requiring continuous visual search for the element being discussed, thereby supporting incremental development of a coherent mental model. In effective presentations, the presenter regulates when the audience is exposed to each element. Progressive disclosure (e.g., through animation or across consecutive slides) allows the presenter to guide attention sequentially and reduce processing burden.
Evidence from eye-tracking studies[8,9] and experimental research on cognitive load, split attention, and redundancy[10,11] further clarifies how different slide designs influence attention:
Decorative or irrelevant images typically attract only brief attention and do not contribute meaningfully to learning.
When slides contain large amounts of text, audiences tend to read the slide, thereby reducing attention to the narration.
When slides contain both dense text and visuals, audiences often engage in frequent gaze switching between elements, increasing cognitive load and potentially impairing comprehension.
When slides present relevant visual representations alongside narration, audiences distribute their attention between the visual content and the narration, shifting between them, which helps the integration of the information.
PROBLEMS WITH AI-GENERATED SLIDES
Slides such as the one in Figure 1 may impair audience attention by violating several of the above principles. They often contain dense bullet lists, redundant text, arrows, labels, infographics, and decorative images, all presented simultaneously. The audience must identify elements, understand relationships, and follow the narration simultaneously. This extraneous load overwhelms working memory[12] and makes it difficult to determine where to allocate attention.
For example, when all elements of a diagram with multiple interacting components are presented at once, the audience must process labels, causal arrows, explanatory text, and narration simultaneously. Some may read ahead, others attempt to follow the speaker, and others focus primarily on the visuals. When attention is directed mainly toward deciphering the slide rather than listening to the speaker, the purpose of a live presentation is undermined.
WORKAROUNDS
A key strategy for improving such slides is progressive disclosure, in which elements are revealed sequentially. This can be achieved in several ways:
Particularly when the elements are spatially separated, overlay each with shapes or cropped images that match the slide background and sequentially remove them using animation. (See the video in the Online Supplementary File 1 for a demonstration. The technique can also be used to reveal cells one by one in tables.)
Create multiple copies of the slide image and remove non-essential portions from each version using image-editing software.
Recreate the slide in PowerPoint using Shapes or SmartArt with concise text, generate relevant images using AI tools,[13] arrange elements to avoid visual clutter, and apply animation.
Another potential approach, as generative AI models evolve, would be to prompt them to sequence complex content across multiple slides, so that related elements are introduced progressively across slides while each slide remains visually complete. However, current models appear limited in their ability to perform such structured segmentation effectively.
In addition, to mitigate the possibility that generative AI outputs may contain inaccuracies or fabricated information, careful human review is essential.[14,15]
Beyond addressing the limitations of AI-generated slides, effective slide design requires attention to several foundational choices—beginning with the template itself. Some general principles for maintaining audience attention are listed in Box 1. These apply regardless of content area and can be implemented without advanced technical skills.
Box 1: General principles for maintaining audience attention
Keep the background and slide design simple and non-distracting. Maintain a consistent style in fonts, sizes, colours, and bullet styles throughout—this will help the audience stay focused on the content.
Use concise language;[16] paragraphs of text invite the audience to read instead of listening. Remove anything that can be inferred from context or is non-essential. Use the slides only as an audience aid; do not use them simultaneously as a speaker prompt and a handout.[1] The narration can be entered in the notes section, which is visible to the speaker during the presentation when Presenter View is used. Prepare separate handouts.
Do not repeat information across text and images or tables.[17]
Wherever possible, replace bullet-point lists with SmartArt, tables, images, or shapes, as monotonous bullet slides disengage audiences quickly.
In the slide title, avoid generic terms such as “Adverse effects”; instead, state the key message (e.g., “The commonest adverse effect is sedation”).[18] Including a section header along with descriptive titles on each slide helps audience members who become temporarily distracted to reorient and re-engage without losing the thread.[1]
When a subtopic spans multiple slides, repeat the slide title with an ellipsis on the continuation slides.
Use section header slides to signal and emphasise transitions between sections.
Distribute handouts only after the presentation, not before, so the audience is not reading ahead during the session.
PROBLEMS WITH UNMODIFIED TEMPLATES
A few templates are bundled with PowerPoint, and numerous others are available for purchase. However, the default font sizes are often too small. Plausible reasons include the possibility that these templates are primarily intended for viewing on personal screens rather than projected displays, that designers aim to make them look information-rich, and that a major user group comprises those preparing marketing reports. (The Design Suggestions feature in PowerPoint, too, often reduces font sizes, likely for the same reasons.) Hence, when using such templates, it is advisable to increase the default font sizes in the Slide Master (to a minimum of ~24–32 pt for body text and ~36–44 pt or larger for titles). Delete irrelevant placeholders within individual slides of a template rather than feeling compelled to fill them.
UTILIZING POWERPOINT’S FEATURES
Features such as slide transitions and WordArt are best avoided, as they may distract from the content.[20] Instead, strategic use of the following PowerPoint features can help control the timing and focus of audience attention:
Guides and gridlines
These alignment tools can help ensure consistent positioning and spacing of elements, improving visual clarity and reducing clutter.
Animation
Revealing successive bullet points one at a time with animation helps keep the audience focused[19] and prevents them from reading ahead. However, this should be done only when there is something substantive to say about each point, as overuse can slow the presentation. Use only the Appear option, as other effects may introduce delay and distraction.
Draw
The pen tool can be used to circle key points for emphasis, draw connections between elements within a complex image, and so on. Using Animation → Replay, these markings can be made to appear sequentially during the presentation [The video in Online Supplementary File 2 demonstrates this].
Highlight
Highlighting can draw attention to key text. Highlighting cannot be animated; however, a similar effect can be achieved by duplicating the slide and highlighting the relevant text in the duplicate. During the presentation, any element can be highlighted in real time (Right-click → Pointer Options → Highlighter).
SmartArt
SmartArt is a useful but underused tool. Online Supplementary File 3 (2.4MB, pdf) illustrates how 75 different SmartArt layouts can be applied in psychiatric presentations.
To improve the effectiveness of SmartArt slides,
Do not use SmartArt when plain text can communicate the idea more clearly. If items exceed ~ 8–10 words, SmartArt often loses clarity.
When choosing a layout—from available layouts or options suggested by the Design Suggestions feature—select one that best emphasizes the underlying relationships rather than prioritizing aesthetics.
Avoid SmartArt styles. Instead, use the cleaner default designs.
In the default color options available for SmartArt, many background colors become progressively lighter across levels, reducing contrast. Manually adjust the background color of shapes with poor contrast.
Left-align long items if sufficient margin is available.
In the Animations tab, select appropriate Effect Options (e.g., One by One, Level at Once) to improve control over information flow.
Shapes
Shapes can be used to convey relationships that SmartArt cannot represent. They can also be used alongside SmartArt on the same slide to add functionality that SmartArt lacks. Arrows and circles, particularly in red, can be used with animation to highlight key points. Speech and thought bubbles are particularly useful for depicting patients’ thoughts or conversations with them. Online Supplementary File 4 (1MB, pdf) provides examples of using different shapes, including text boxes, in slides.
Many default Shape Styles lack sufficient contrast for text. Avoid gradient backgrounds, outlines (especially in darker colors), and Shape Effects.
Morph animation
While SmartArt is well-suited to represent relationships, Morph animation is the better option to represent processes. It is particularly useful when the goal is to convey dynamic change, temporal flow, or transformation of states. Examples include:
Demonstrating evolution of symptoms over time (e.g., suspiciousness → fixed delusion)
Explaining neurobiological mechanisms (e.g., normal → altered neurotransmission)
Presenting stepwise clinical reasoning or formulation (presenting symptoms → differential diagnosis → final diagnosis)
Depicting dose–response or risk-factor relationships through changes in size or magnitude
Conveying conceptual shifts, such as moving from categorical (DSM) to dimensional or transdiagnostic models, where the same symptom clusters reorganise on screen.
See the video in the Online Supplementary File 5 for a demonstration. To use Morph effectively, layouts should remain consistent across slides. Keeping narration minimal allows the motion itself to carry part of the explanatory load.
Section Zoom
This feature helps break a long presentation into clearly bounded sections, effectively creating mental “chapters” that allow the audience to reset and re-engage at each transition to another section. An overview slide functions as a persistent anchor, indicating both where the presentation currently is and what lies ahead, thereby reducing disorientation and freeing cognitive resources for content processing rather than navigation. Returning to the overview slide (e.g., by typing the slide number and pressing Enter) re-establishes the broader framework and reinforces how the material just covered fits within the whole.
This approach is particularly useful in case-based discussions, where one may move flexibly between sections such as history, mental status, formulation, and management, or begin with a high-level biopsychosocial model and selectively zoom into biological, psychological, or social domains as required. See the video in the Online Supplementary File 6 for a demonstration.
Accessibility Assistant
Available at Review → Check Accessibility, this tool helps discover issues such as text with poor contrast, decorative shapes, and missing slide titles.
Pointer
When a laser pointer is not available or during online presentations, the mouse can function as one (Right-click in the Slide Show mode and select Pointer Options → Laser Pointer).
Zoom
This feature can be used to enlarge and display a specific portion of a complex image. The source image should have sufficient resolution to remain clear when magnified. Free AI-based image upscalers can help enhance resolution if needed.
EFFECTIVE DATA VISUALIZATION
Use the most appropriate type of graph to convey the data[20,21]; for example, do not use pie charts to convey changes over time. Avoid 3D graphs, as the 3D effect can distort data interpretation and increase cognitive load without adding informational value. Avoid non-essential visual elements (e.g., heavy gridlines, borders, background shading, decorative effects) to direct attention to the data.[21] Font sizes in graphs exported from statistical software are often too small for projected display and would need to be enlarged before embedding in slides. Avoid conveying information solely through color; use labels, patterns, or annotations to ensure accessibility for viewers with color vision deficiencies.
Instead of using screenshots of tables from journal articles, recreate them in PowerPoint. Include only the relevant rows and columns. AI tools can be used to extract text from screenshots; however, the output should be verified for accuracy (text from screenshots of article content is also better extracted and presented in this manner).
AI TOOLS WITHIN POWERPOINT
Recently, AI tools that integrate natively into PowerPoint (as add-ins that work within the application) have become available. Examples include Copilot (available with Microsoft 365 subscriptions), Claude by Anthropic for PowerPoint, Plus AI for PowerPoint, and Auxi AI.
Relevant features in these tools include generating coherent, well-structured outlines; compressing long text into concise bullet points; arranging text, images, and spacing for clarity; suggesting, and in some cases generating, context-relevant visuals for each slide; and ensuring design consistency across slides. For example, feedback from the inbuilt “/ib-check-deck” prompt in Claude by Anthropic for PowerPoint included, “A CrossMark “Check for updates” badge from a journal screenshot has been left on the slide—a web artefact with no purpose in a presentation.”
However, human oversight remains essential to ensure accuracy and cultural appropriateness. Besides, the default outputs of these tools appear to be optimized for self-contained, on-screen viewing rather than narrated presentations. Therefore, when prompting such tools, explicitly specify that the content is intended for a narrated presentation. Defining the target audience (e.g., psychiatry postgraduate students) can further improve relevance and focus.
SLIDES FOR ONLINE PRESENTATIONS
As online viewers often use smaller screens or split windows, larger font sizes (≥28–32 pt for body text), and high contrast (dark text on light backgrounds or vice versa) are recommended. Because audio–visual mismatch is common in online settings, slides should be partially self-explanatory and include key terms and critical content that might otherwise be conveyed verbally. Avoid placing key content in areas where video thumbnails may overlap. Heavy animations can appear choppy or lag during screen sharing, particularly over unstable connections; simpler entrance effects are more reliable in online contexts. Testing slides on smaller devices and under reduced brightness can help ensure adequate visibility for diverse viewers. Unlike in-person settings, disengagement is less visible online; therefore, incorporating interactive elements such as polls and questions at periodic intervals is advisable.[22,23]
USING NARRATION CAREFULLY
Narration is more important than slides, which serve only as a supplement and visual aid. Audiences have different learning preferences: while some may be better served by a visually rich presentation, others may prefer to learn aurally or through reading. Ideally, a presentation should cater to both groups.[1] Reserve some content for narration rather than placing everything on the slides.
Avoid reading from slides verbatim, as it reduces engagement. Conversely, interacting with the audience and adopting a conversational delivery style enhance engagement.[1] Advance to the next bullet or slide slightly before finishing the explanation of the current one to maintain a smooth flow of attention. When moving to a slide with a complex image or graph, pause briefly to allow the audience time to process it. During the presentation, the “B” key can be used to black out the screen when attention needs to be directed to the speaker rather than the slide, for example, during extended verbal explanations or discussions.
Two PowerPoint features that help improve narration are:
Rehearse with Coach
Making a mock presentation using this feature provides feedback on the pace of speech, the use of filler words, excessive repetition of the same word or phrase, awkward pauses, overly long silences, and a lack of variation in tone and pitch; it also suggests more concise or precise wording.
Presenter View
During the presentation, Presenter View can assist in multiple ways, including displaying a preview of the next slide, thereby facilitating smoother narrative transitions.
CONCLUSION
Effective slide design is fundamentally the deliberate control of when and where the audience attends. A common underlying mechanism in poorly designed slides is increased extraneous cognitive load, which leads to misallocation of limited attentional resources and reduced integration of information. Judicious use of PowerPoint features and AI tools, with consideration of the audience context, can guide attention rather than bombard the audience with distractions, helping the message stand out in an era of widespread inappropriate slide design and reduced attention spans. These suggestions should be interpreted considering their limitations: although grounded in established theory, most have not been empirically validated, and they reflect, to some extent, the author’s judgement.
Conflicts of interest
There are no conflicts of interest.
Video available on: https://journals.lww.com/indianjpsychiatry
Acknowledgements
Generative artificial intelligence tools, ChatGPT (OpenAI) and Claude (Anthropic), were used for language refinement and to assist in identifying some of the cited literature. All outputs were critically reviewed and edited by the author, who takes full responsibility for the content of the manuscript. Figure 1 was created using ChatGPT (OpenAI).
Funding Statement
Nil.
REFERENCES
- 1.Undrill G, McMaster F. PowerPoint: Avoiding the slide to damnation. Adv Psychiatr Treat. 2013;19:14–22. [Google Scholar]
- 2.Mona AE, Roshith V, Peter R, Roy P, Hassan A, Devika M, et al. Short video addiction and its impact on cognitive functioning in adolescents and youth: A systematic review. Int J Adolesc Youth. 2026;31:e2623337. [Google Scholar]
- 3.Xie J, Xu X, Zhang Y, Tan Y, Wu D, Shi M, et al. The effect of short-form video addiction on undergraduates’ academic procrastination: A moderated mediation model. Front Psychol. 2023;14:1298361. doi: 10.3389/fpsyg.2023.1298361. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Clark RC, Mayer RE. Hoboken (NJ): John Wiley and Sons; 2016. E-Learning and the Science of Instruction: Proven Guidelines for Consumers and Designers of Multimedia Learning. [Google Scholar]
- 5.Sweller J, Ayres P, Kalyuga S. New York: Springer; 2011. Cognitive load theory. [Google Scholar]
- 6.Mayer RE. 2nd. New York: Cambridge University Press; 2009. Multimedia Learning. [Google Scholar]
- 7.Mayer RE, Moreno R. Nine ways to reduce cognitive load in multimedia learning. Educ Psychol. 2003;38:43–52. [Google Scholar]
- 8.Hegarty M, Just MA. Constructing mental models from text and diagrams: Evidence from eye movements. J Mem Lang. 1993;32:717–42. [Google Scholar]
- 9.Schmidt-Weigand F, Kohnert A, Glowalla U. A closer look at split attention in multimedia learning: Evidence from eye-tracking. Learn Instr. 2010;20:100–10. [Google Scholar]
- 10.Mayer RE, Heiser J, Lonn S. Cognitive constraints on multimedia learning: When presenting more material results in less understanding. J Educ Psychol. 2001;93:187–98. [Google Scholar]
- 11.Kalyuga S, Chandler P, Sweller J. Managing split attention and redundancy in multimedia instruction. Appl Cogn Psychol. 1999;13:351–71. [Google Scholar]
- 12.Miller GA. The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychol Rev. 1956;63:81–97. [PubMed] [Google Scholar]
- 13.Ameen S. Beyond Google Images: Crafting impactful images for presentations or clinical use with AI. Indian J Psychiatry. 2025;67:1016–20. doi: 10.4103/indianjpsychiatry_971_25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Salvagno M, Taccone FS, Gerli AG. Can artificial intelligence help for scientific writing? Crit Care. 2023;27:75. doi: 10.1186/s13054-023-04380-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Athaluri SA, Manthena SV, Kesapragada VSRKM, Yarlagadda V, Dave T, Duddumpudi RTS. Exploring the boundaries of reality: Investigating the phenomenon of artificial intelligence hallucination in scientific writing through ChatGPT references. Cureus. 2023;15:e37432. doi: 10.7759/cureus.37432. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Apperson JM, Laws EL, Scepansky JA. An assessment of student preferences for PowerPoint presentation structure in undergraduate courses. Comput Educ. 2008;50:148–53. [Google Scholar]
- 17.Ameen S, Praharaj SK, Menon V. “Two minutes more!” Preparing slides for conference research presentations. Indian J Psychol Med. 2023;45:1–4. doi: 10.1177/02537176221142555. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Alley M, Schreiber M, Ramsdell K, Muffo J. How the design of headlines in presentation slides affects audience retention. Tech Commun. 2006;53:225–34. [Google Scholar]
- 19.Jones AM. The use and abuse of PowerPoint in teaching and learning in the life sciences: A personal overview. Biosci Educ. 2003;2:1–13. [Google Scholar]
- 20.Midway SR. Principles of effective data visualization. Patterns (N Y) 2020;1:100141. doi: 10.1016/j.patter.2020.100141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Tufte ER. 2nd. Cheshire (CT): Graphics Press; 2001. The Visual Display of Quantitative Information. [Google Scholar]
- 22.Means B, Toyama Y, Murphy R, Bakia M, Jones K. Washington (DC): US Department of Education; 2009. Evaluation of Evidence-Based Practices in Online Learning: A Meta-Analysis and Review of Online Learning Studies. [Google Scholar]
- 23.Dhawan S. Online learning: A panacea in the time of COVID-19 crisis. J Educ Technol Syst. 2020;49:5–22. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.