🎓 Text Doc Digital Accessibility Converter

📋 Applying Heading Styles in Word

Heading styles (Heading 1, Heading 2, Heading 3, Title) are different from heading formatting (bold, larger font). A paragraph that looks like a heading because it is bold and large is invisible to screen readers — it is just bold text. Only a paragraph with a proper heading style applied creates a navigable outline.

How to apply a heading style:

1. Click anywhere in the paragraph you want to make a heading.
2. In the Home ribbon, find the Styles gallery (a row of style thumbnails). Click Heading 1, Heading 2, or Heading 3 as appropriate.
3. If you don't see the Styles gallery, click the small arrow at the bottom-right of the Styles group to open the Styles pane.
4. For the document title, use the Title style (also in the Styles gallery).

Tip: You can still apply bold, italic, or a different font size on top of a heading style without losing its accessibility. The style is what matters to screen readers, not the visual appearance.

Hierarchy: Use Heading 1 for top-level sections, Heading 2 for sub-sections within those, and Heading 3 for sub-sub-sections. Do not skip levels (e.g., do not jump from Heading 1 directly to Heading 3).

🗂️ Why Heading Structure Matters — and How to Navigate It

Heading styles do more than make text look like a heading. They create an invisible structural outline of your document that both sighted users and screen reader users depend on to find their way around. Think of it as a built-in, always-current table of contents embedded in the document's structure.

For screen reader users: Screen readers can present a list of all headings in a document, letting users jump directly to any section without reading through everything in between. This is often the primary way blind or low-vision users navigate long documents — the same way a sighted reader skims headers to find the section they want. Without proper heading styles, that navigation simply does not exist: the document becomes a single undifferentiated block of text.

For all users — the Navigation Pane: Word's Navigation Pane is one of its most powerful and least-known features. It displays your document's heading structure as a clickable outline on the left side of the screen, letting you jump to any section instantly and see the shape of your document at a glance. For anyone writing long documents — lecture notes, reports, course materials — it is genuinely transformative.

How to open the Navigation Pane:

1. Go to the View tab in the ribbon.
2. Check the Navigation Pane checkbox in the Show group.
3. The pane appears on the left. Click the Headings tab at the top of the pane (the other tabs are Pages and Results).
4. You will see your document's heading structure as a collapsible outline. Click any heading to jump directly to that section.

If your Navigation Pane shows nothing under Headings, it means your document has no paragraphs with real heading styles applied — which is exactly what this converter helps you fix.

How to use the outline effectively:

• You can drag headings in the Navigation Pane to reorder entire sections of your document — the heading and all its content move together.
• You can right-click a heading in the pane to promote or demote its level, delete the section, or add a new heading before or after it.
• The outline gives you an immediate visual check on your document's structure — if it looks wrong in the Navigation Pane, it will feel wrong to a reader navigating with a screen reader too.

The accessibility connection: A well-structured heading outline is the single most impactful accessibility improvement you can make to a long document. It benefits screen reader users, users of other assistive technologies, users who rely on keyboard navigation, and — as the Navigation Pane demonstrates — every other reader as well. Accessibility and good document structure are the same thing.

Carrying heading structure into PDF: When you export a Word document to PDF, you can embed the heading structure directly into the PDF so it remains navigable for screen reader users — not just in Word. In Word: File → Export → Create PDF/XPS, then click Options… and check "Document structure tags for accessibility" before clicking Publish. The resulting PDF will have a proper heading outline, bookmarks panel, and reading order that assistive technologies can use — but only if the headings were properly styled in the Word document first.

🎨 Modifying Heading Styles Globally in Word

When you apply a heading style, Word uses its built-in appearance — often a blue, serif font that may not match your document's design. You do not need to reformat every heading individually. Instead, redefine the style itself once, and every paragraph using that style updates automatically throughout the entire document.

How to modify a heading style globally:

1. Apply the heading style to one paragraph (e.g., Heading 1) as described above.
2. Manually format that one paragraph the way you want all Heading 1s to look — font, size, color, bold, spacing, etc.
3. Right-click Heading 1 in the Styles gallery on the Home ribbon.
4. Select "Update Heading 1 to Match Selection".
5. All paragraphs styled as Heading 1 throughout the document instantly update to match your new formatting.

Alternative — edit the style directly:

1. Right-click Heading 1 in the Styles gallery → Modify…
2. Set font, size, color, and spacing in the dialog. Click Format at the bottom for additional options (paragraph spacing, borders, etc.).
3. Make sure "New documents based on this template" is selected if you want your changes to carry over to future documents. Click OK.

Tip: Repeat this for Heading 2 and Heading 3 to create a consistent, custom heading hierarchy that is both visually on-brand and fully accessible. The accessibility is carried by the style name, not the visual appearance — so you can make headings look however you like without sacrificing screen-reader compatibility.

🖼️ Adding Alt Text to Images in Word

Alt text (alternative text) is a written description of an image that screen readers read aloud to users who cannot see the image. Without it, screen reader users hear either nothing or the raw file name (e.g., "image001.png"), which is useless.

How to add alt text:

1. Right-click the image.
2. Select Edit Alt Text… (or Format Picture → Alt Text on older versions of Word).
3. Type a description in the text box. Click outside when done.

What makes good alt text:

• Describe what the image shows and why it matters in context — not just what type of image it is.
• Do not start with "Image of…" or "Photo of…" — screen readers already announce it as an image.
• Keep it concise: one or two sentences is usually enough.
• For graphs and charts, describe the key finding or trend, not every data point.
• For purely decorative images (dividers, backgrounds), check the Mark as decorative checkbox instead of writing alt text.

Example: Instead of "Graph", write "Bar graph showing bacterial colony counts declining by 80% over 24 hours following phage treatment."

Using an AI to generate alt text: AI assistants (ChatGPT, Claude, Gemini, etc.) can help write alt text descriptions. Paste the image into the AI chat along with this prompt:

"Please write a concise alt text description for this image suitable for use with a screen reader. Describe what the image shows and why it matters in an academic or educational context. Keep it to 1-2 sentences. Do not start with 'Image of' or 'Photo of'. Focus on content and significance."

The Review & Fix tab includes Copy AI prompt and Copy image buttons next to each image to make this workflow easier.

Surrounding text. Each image also shows the document's own words around it — the caption or paragraph the image sits in, the text before it, and the text after it. This is usually the fastest way to work out what the alt text needs to convey, because the surrounding prose normally explains what the figure is for. A Copy context button copies that text if you want it elsewhere, and the Copy AI prompt button now includes it automatically, so an AI assistant can see how the image is used rather than guessing from pixels alone. Describe what is actually visible in the image; the context tells you which parts of it matter.

Alt text that is not really alt text. An image counts as lacking a description when the field is empty or when what is in it describes nothing: a file name, a full path, a web address, a bare label such as “Picture 3”, or a filename stem like zebra-cu-half. Word treats all of those as filled in and will not complain, but read aloud they tell a listener nothing, and accessibility checkers reject them. The card says which case it found and shows you the offending text, so it is clear the tool has not simply overlooked what was there. Genuine descriptions are never touched, including ones that happen to mention an address, such as “Diagram of the lytic cycle, see phage.org for detail”.

Color accessibility: Alt text addresses screen reader accessibility, but images also need to be accessible to colorblind readers. If your figures use color to convey information, check them at colors.phage.org — a free tool that simulates how images appear under different types of color vision deficiency. The Review & Fix tab includes a Download image button so you can save a figure, adjust its colors in the Colors tool, and use Replace image to swap the revised version back into the document.

📊 Tables and Digital Accessibility — A Practical Guide

Tables are one of the most challenging accessibility problems in Word documents, especially those converted from PDF. This guide covers what the issues are, what works, what doesn't, and what you can do instead.

Why table accessibility matters: Screen readers navigate tables by reading column and row headers aloud as the user moves between cells. Without designated headers, every cell sounds the same — the user has no way to know which column or row they are in. A table without headers is effectively unnavigable for screen reader users.

The standard fix — and why it may not work

The standard approach is:

1. Click anywhere in the first row of your table.
2. Go to the Table Design tab → check Header Row.
3. Right-click the first row → Table Properties → Row tab → check "Repeat as header row at the top of each page". Click OK.

Known Word limitation: In Word 365 and some recent versions, it is not possible to designate only the first row as a header row — the option either does not apply correctly, or marks all rows as headers. This appears to be a bug in Word. Additionally, right-clicking to access Table Properties may not be available in all versions. If this is the case, do not spend time trying to force Word to comply — use one of the alternative approaches below.

Tables converted from PDF — a different problem

When a PDF is converted to DOCX, tables rarely survive as real Word tables. They typically become either plain text with manual spacing, or a grid of text boxes. In these cases, "Convert Text to Table" in Word usually does not work because the text structure does not match what Word expects. Do not waste time trying to reconstruct the table structure manually.

Alternative approach 1 — Convert the table to text

If the table content can be understood as a list or sequence of text, converting it to plain text is often the cleanest accessible solution:

1. In Word: select the table → go to Table Design or Layout tab → click Convert to Text.
2. Choose a separator (tabs work well for columnar data; paragraph marks work for simple lists).
3. The result is plain text that screen readers handle perfectly.

This converter's Review & Fix tab shows a tab-separated text preview of each detected table, with a copy button. You can paste this directly back into your Word document as a text alternative to the table.

Alternative approach 2 — Convert the table to an image

If the table's visual layout is important (complex headers, merged cells, colour coding), converting it to an image with a thorough alt text description is a valid approach:

1. In Adobe Acrobat: use Edit PDF → Snapshot tool to capture the table as an image.
2. In Word: take a screenshot of the table, or use Insert → Screenshot.
3. Insert the image into your document and write a comprehensive alt text description covering the table's content and key findings.

Note: This approach turns a structural accessibility problem into an alt text problem — which is solvable. A well-written alt text description of a table can fully convey its content to a screen reader user.

Alternative approach 3 — Distribute DOCX instead of PDF

Untagged PDFs (those converted without accessibility structure) cannot be made fully accessible after the fact. If your document is primarily text with tables, distributing the DOCX file directly — rather than exporting to PDF — sidesteps the problem entirely. DOCX files are natively accessible to Word's own accessibility tools and screen readers.

If you must distribute a PDF, export it from Word using File → Export → Create PDF/XPS → Options → Document structure tags for accessibility. This embeds heading and structure tags into the PDF, making it navigable. However, table header accessibility in the resulting PDF still depends on whether Word correctly marked the headers beforehand.

Tables that span multiple pages

For tables that span multiple pages, the header row needs to repeat at the top of each page. If Word's "Repeat as header row" option is working in your version, use it. If not, the table-to-text or table-to-image approaches above are the most reliable alternatives.

Not every grid is a data table

A header row is only meaningful when other rows depend on it. “12” sitting under a column headed “Plaques” means something because of the header; a box saying “Draw here” heads nothing. Tables used purely to arrange things on the page — drawing boxes, side-by-side prompts, a grid of labels — are layout tables, and marking a header row on one is not merely pointless but wrong, because it tells a screen reader there is a relationship to announce when there is not. Accessibility checkers make the same distinction, which is why a lab sheet full of drawing boxes can pass without a single header row while a table of results still fails.

The converter therefore judges each table before proposing anything, and says what it concluded and why. It treats a table as needing no header when it has only one row or one column, when every cell is a prompt ending in a colon, or when no row stands out from the rest. It treats one as a data table when the first row is bold or shaded, when the first row is much shorter than the rows below, or when the rows below hold numbers and the first row does not. Where it cannot tell, it says so and leaves the decision to you. The header fix is only ticked in advance for tables that look like data.

Summary — what actually works

• ✅ Remove the table, with the vertical space preserved — Review & Fix offers three independent, combinable table operations. Tick 1. Remove this table and the w:tbl element is deleted, so a checker no longer sees a table at all. 2. Preserve the vertical space then inserts an empty paragraph of the same height, so surrounding text and floating images do not shift. Where every row carries an explicit height the spacer is exact; where rows are auto-height it is estimated and you can type the height in inches yourself.
• ✅ Insert a picture of the table, with alt text — 3. Insert a picture of the table draws the table to an image using its real column widths, embeds it where the table was, and writes your alt text into the picture's description. Combine it with removal to keep the content visible while presenting it as a described image rather than a table.
• ✅ Convert floating images to inline — often the cleanest fix of all. A floating image (Word's "in front of text" / "behind text" wrapping) sits outside the text flow: it reserves no vertical space, so text can run underneath it, and screen readers frequently skip it or announce it out of order. The Images group in Review & Fix offers Convert floating images to inline, which places every such image into the reading order and gives it real space on the page, pushing following text below it. If your tables existed only to hold space open beneath a floating image, converting to inline removes the need for spacers entirely — so ticking this box automatically switches off the per-table spacer to avoid counting the space twice.
• ℹ️ Images inside a removed table are kept — if a table contains pictures, they are moved out and placed above the replacement rather than deleted with the table, and are made inline so they hold their own space. Any spacer height is reduced by the space those images now occupy. The converter warns you when a table contains images.
• ✅ Convert table to text — most reliable; fully accessible everywhere; use the tab-separated text preview in Review & Fix
• ✅ Convert table to image + alt text — reliable for visual layouts; requires a thorough description
• ✅ Distribute DOCX instead of PDF — screen readers handle DOCX tables better than untagged PDFs; sidesteps the checker problem
• ⚠ "Fix header rows" in this converter — writes correct OOXML markup; may help real screen readers and external checkers even if Word's own checker doesn't acknowledge it; worth applying alongside text/image alternatives
• ❌ Word's Header Row checkbox — broken in Word 365; do not rely on it
• ❌ Reconstructing PDF tables in Word — generally not worth attempting

📋 Creating Proper Lists in Word

Typing a bullet character (•) or a number (1.) manually and pressing Tab creates text that looks like a list but is not recognized as one by screen readers. Use Word's built-in list formatting instead.

How to create a proper list:

1. Place your cursor where you want the list to begin.
2. In the Home ribbon, click the Bullets button (●≡) for a bulleted list, or the Numbering button (1≡) for a numbered list.
3. Type your list items, pressing Enter between each one.
4. Press Enter twice (or click the button again) to end the list.

To convert existing manual list items: Select the paragraphs, then click the Bullets or Numbering button. Word will apply proper list style and you can delete the manual characters.

✅ Checking Accessibility in Word

Word has a built-in accessibility checker that complements this converter. It can catch issues this tool may miss, and will confirm that the fixes applied here have taken effect in the final document.

How to run the accessibility checker:

1. Go to the Review tab in the ribbon.
2. Click Check Accessibility.
3. The Accessibility pane opens on the right, listing errors, warnings, and tips.
4. Click any item in the list to jump to the relevant element in the document and see suggested fixes.

Common findings and what they mean:

• "Missing Alt Text" — an image has no description. Right-click → Edit Alt Text.
• "Table has no header" — mark the first row as a header row (see above).
• "Document has no title" — apply the Title style to your document title, or set the document title in File → Properties.
• "Hard-to-read text contrast" — text color is too close to the background color for low-vision readers.

Running Word's checker after downloading your corrected file from this converter is a good final step before distributing it.

📝 Using Google Docs Instead of Word

If you author documents in Google Docs rather than Word, most of the same accessibility principles apply. Google Docs supports heading styles, alt text on images, and document outline navigation — and documents can be exported to .docx for use with this converter.

Applying heading styles in Google Docs:

1. Click anywhere in the paragraph you want to make a heading.
2. Open the Format menu → Paragraph styles → choose Title, Heading 1, Heading 2, or Heading 3.
3. Alternatively, use the style dropdown in the toolbar (it usually shows "Normal text") and select the heading level from there.

Modifying heading styles globally in Google Docs:

1. Format one paragraph the way you want all headings of that level to look.
2. Go to Format → Paragraph styles → Heading 1 (or whichever level) → "Update 'Heading 1' to match".
3. All paragraphs of that heading level update throughout the document.

The Document Outline — Google Docs' Navigation Pane equivalent:

1. Go to View → Show document outline (or press Ctrl+Alt+A / ⌘+Alt+A).
2. The outline panel appears on the left, showing all headings as clickable links.
3. Click any heading to jump to that section instantly.

As with Word, if the outline is empty, it means no real heading styles have been applied.

Adding alt text to images in Google Docs:

1. Right-click the image → Alt text.
2. Fill in the Description field. (The Title field is optional.)
3. Click OK.

Table headers in Google Docs:

This is where Google Docs falls short of Word. As of 2025, Google Docs does not provide a built-in way to mark a table row as a header row in the accessibility sense that screen readers recognize. The visual appearance of a header row can be achieved with formatting (bold, background color), but this does not create accessible header markup. If accessible tables are important for your document, consider completing the table in Word or using this converter's output as a starting point and finishing in Word.

Exporting to .docx for use with this converter:

1. In Google Docs: File → Download → Microsoft Word (.docx).
2. Upload the downloaded .docx to this converter.
3. Note that heading styles you applied in Google Docs will be preserved in the exported .docx, so the converter will recognize them as real headings — a good reason to apply styles in Google Docs before exporting.

Checking accessibility in Google Docs: Google Docs does not have a dedicated accessibility checker equivalent to Word's. The best approach is to export to .docx, run it through this converter, and then use Word's built-in accessibility checker on the result.

🔗 Hyperlinks and Plain-text URLs

There are three categories of link issues this tool detects:

• Non-descriptive hyperlinks in body text — clickable links whose display text is a bare URL or generic phrase like "click here". These need descriptive replacement text that makes sense out of context: "2023 phage therapy review in Lancet", not a bare URL.
• In-text citation links — hyperlinks with display text like [1] or [2], linking internally to the reference list. These are not an accessibility problem — they are a well-understood academic convention. The tool shows them collapsed for information only.
• Reference-section links — URLs embedded in citations after the References heading. DOI and PubMed URLs are flagged as probably-deletable; URLs with "accessed/retrieved" language nearby are flagged as probably-essential (the URL is the reference itself). The section is collapsed by default since reference links are often acceptable as-is.

Plain-text URLs appear as unclickable text — often because PDF conversion strips hyperlink formatting. The converter detects these and offers "Make live" to convert them to clickable hyperlinks with optional descriptive display text.

Tip: In Word, select any plain-text URL and press Ctrl+K (Cmd+K on Mac) to make it a live hyperlink manually.

🖥 PowerPoint Quick Fix

The Slides tab is deliberately limited. It makes two blunt changes to a .pptx file and reports a few things it cannot fix. It exists for decks that are kept online for reference — archived lectures where the teaching content is really in the videos — rather than material students are expected to work from directly.

The changes are independent tick boxes, so you can apply any combination: white backgrounds, black text, clear coloured backgrounds behind text, and alt text from slide titles. A fifth sets the presentation title in the file properties if it is missing.

Random Effects entrance animation

PowerPoint once offered an entrance effect called Random Effects, which chose a different animation each time a slide was shown. The option was removed from the interface, but the file format and the player still support it, so a deck that has it keeps working — it simply cannot be applied through PowerPoint any more. This tab can apply it.

Tick Give images a “Random Effects” entrance animation and every picture is listed so you can untick any that should stay still.

The list distinguishes two cases, and defaults differently for each. An image that already has an entrance effect starts ticked: its effect is converted to Random Effects in place, keeping its position in the click order, so the slide advances exactly as it did before. An image with no animation starts unticked, because adding animation where there was none is a change to how the slide behaves rather than a change to how an existing animation looks — tick it and it gains an entrance of its own on a new click. Nothing is animated that you did not ask for.

Each entry shows a thumbnail of the picture alongside its slide number and title, so you can tell at a glance which image you are deciding about. Presentations sometimes hold pictures in vector formats that browsers cannot draw, such as EMF or WMF; those entries show a placeholder instead of a preview, and are otherwise treated identically.

Each entry says which case it is and, where the file records it, what the current effect is. An image already set to Random Effects is left alone, so running this twice changes nothing the second time. Animations belonging to text and other shapes are never altered.

The markup written is exactly what PowerPoint itself produces — presetID="24" presetClass="entr" presetSubtype="0" with a placeholder behaviour carrying an empty attribute name, which is how the format signals that the effect is left to chance.

The downloaded file is named for what was actually applied: _accessible when any of the accessibility options ran, _random when the animation ran, and _accessible_random when both did. Untick everything under What to change and tick only the animation, and you get a file suffixed just _random — useful when you want the animation without altering how the deck looks. Tick all and Untick all apply to the accessibility options only; the animation is left as you set it.

This one is filed under presentation polish rather than accessibility, and deliberately so. Click-triggered animations hide content until someone interacts, which assistive technology may never do, and motion causes difficulty for people with vestibular disorders. It is off by default.

Animations are left alone. Slide timing lives in a separate part of each slide's XML, and none of these changes touch it. A deck carrying entrance effects comes through with every animation intact — verified on a nineteen-slide deck with eighty-eight animation nodes, which came out with all eighty-eight unchanged.

A note on old .ppt files. This tab reads .pptx only. A file still in the old binary .ppt format will be refused, and PowerPoint will also have it in Compatibility Mode, which is the usual reason Review → Check Accessibility appears greyed out. Use File → Info → Convert, or Save As with the type set to PowerPoint Presentation, and both problems go away together.

Backgrounds white, text black. A white background is written onto every slide, which overrides whatever the layout and slide master specify, and every text run is set to black. Shape fills too dark to read black text against are cleared, judged by the same 4.5:1 contrast rule used elsewhere in this tool, with theme colours resolved from the presentation's own colour scheme. Text sitting on top of a photograph cannot be fixed this way, so those slides are reported for you to look at yourself.

Alt text from slide titles. Pictures with no usable alt text are given “Image for [slide title]”. Note what this is and is not: it satisfies an automated checker, and it tells a screen reader user which slide they are on, but it does not describe the picture. Where an image carries real content, replace it with a proper description. Existing alt text is never overwritten — though PowerPoint's habit of pre-filling the field with the image's filename counts as no alt text, since “IMG_2841.JPG” read aloud is worse than useless.

Alt text that is really a filename. PowerPoint frequently pre-fills the alt text field with whatever it can find — often the image's filename, and sometimes the entire path it was inserted from, such as C:\Users\you\Documents\Figures, landscape (011).jpg. Accessibility checkers reject these, and read aloud they are worse than silence. Anything ending in an image extension, anything beginning with a drive letter or UNC path, and bare labels like “Picture 3” all count as missing alt text and will be replaced. So do two subtler forms: text carrying percent-escapes, which only ever come from a filename or URL (Zebra%20Row50), and single tokens joined by hyphens or underscores with no spaces at all (zebra-cu-half, fig_3_final), which are filename stems rather than descriptions. Real descriptions are left alone — including hyphenated ones like “Phage-host interaction over time”, since the presence of spaces marks it as prose.

Text over pictures. Text laid directly on a photograph may be unreadable once it is blackened, so those shapes are reported rather than changed. Callouts, labels and coloured title bars are not counted, because their own fill keeps the text readable — only bare, unfilled text boxes sitting over an image are flagged.

Hyperlink colours. These are the one kind of text colour that making everything black does not reach, because PowerPoint takes them from the theme's colour scheme rather than from the text itself — so a deck can have perfectly black body text and still fail on its links. The theme entries are hlink for links and folHlink for visited links, and a deck may carry more than one theme part, all of which are checked.

Failing colours are replaced with Office's own defaults, #0563C1 and #954F72. Those were chosen on the numbers rather than habit: each clears 4.5:1 against a white background, and each also clears 3:1 against black body text, so a link remains distinguishable from the text around it and not only by being a different colour. Nothing else in the colour scheme is touched.

Coloured backgrounds behind text. Fills are judged by the same 4.5:1 rule. A fill too dark for black text is always cleared, since blackening the text on it would make matters worse. A fill that black text reads well against — a yellow title bar is about 19:1 — is left alone by default, because there is nothing wrong with it. If you would rather everything sat on plain white, tick Also clear coloured backgrounds behind text; that is a decision about appearance, not accessibility, which is why it is off by default.

What it does not do. It cannot invent slide titles, and a slide with no title placeholder is the single most common PowerPoint accessibility failure — screen reader users move between slides by title. Add them in PowerPoint under Home → Layout or in the Outline view. It also does not touch reading order, the other common problem: check that under Home → Arrange → Selection Pane, where shapes are announced from the bottom of the list upwards.

⛓ Removing Hyperlinks

Worth being clear that this is document cleanup, not an accessibility fix. A hyperlink is not a fault. What accessibility standards ask is that link text describes where the link goes, which is handled by the Non-descriptive Hyperlinks section. Stripping links will raise a checker score, but by removing what was being judged rather than by improving it, and it takes away function for every reader including the ones the standards exist to help.

If your aim is to satisfy a checker while keeping the links working, the alternatives are: rewrite the link text so it describes the destination; or, for a document that will be both printed and read on screen, leave the URL visible and make the sentence around it descriptive — WCAG 2.4.4 allows a link's purpose to come from its context, so “Watch the introductory video: https://…” is acceptable.

It is offered because removing links is genuinely useful for other reasons: preparing a version for print, clearing automatically linked email addresses, or stripping links to resources that have gone.

What it does. The link is removed and the words stay. The Hyperlink character style goes with it, so nothing is left blue and underlined and still looking clickable when it is not. Links whose visible text is itself a URL or an email address are excluded by default, since unlinking those leaves a dead web address sitting in the text — a problem this tool flags elsewhere in its own right; tick Include the links whose text is itself a URL if you want them too. Cross-references and tables of contents are never touched, because those are navigation, and removing them would take an accessibility feature away rather than add one.

Word's own equivalent, selecting all and pressing Ctrl+Shift+F9, does not make that distinction: it flattens every field in the document, so cross-references, page numbers and any table of contents become plain text that no longer updates.

👁 Hidden Text and 💬 Comments

Hidden text is text formatted with Word's Hidden font attribute (Format → Font → Hidden). It does not display or print, but it is still in the file: it travels with the document when you share it, text extraction can reach it, and it may be exposed by other tools. Authors often use it for drafting notes they never intend to publish.

The converter lists each hidden passage so you can delete them individually or all at once. Passages inside tracked changes, field codes, footnotes or comment references are deliberately not offered for deletion — removing those runs damages the surrounding structure rather than tidying the document.

To see hidden text in Word, turn on the paragraph marks button (¶) on the Home tab; hidden text then appears with a dotted underline.

Comments are review annotations. They are not an accessibility fault in themselves, but they stay in the shared file, some screen readers announce them, and they are usually not meant for the final reader. Deleting a comment here removes its text from comments.xml along with its anchor marks and its reference in the document body, so nothing is left dangling.

🔄 Using This Tool Iteratively

This tool works best across multiple passes rather than a single session:

1. First pass — upload the original document, review the Critique, accept the clear fixes (language, obvious headings, lists). Skip anything uncertain.
2. Download and re-upload — the corrected DOCX becomes the input for the next pass. Issues marked Skip will reappear; issues marked Ignore will not.
3. Subsequent passes — address remaining issues with more context. Headings you previously rejected are tagged with a NotAHeading style and will not be re-flagged.
4. Final verification — run the finished document through Word's built-in checker (Review → Check Accessibility) and/or your institution's designated accessibility checker to confirm.

Remember: there is a difference between digitally accessible, aesthetically acceptable, and perfect. Achieving two out of three is sufficient for most applications — do not let the perfect be the enemy of the accessible!

About this tool

The Text Doc Digital Accessibility Converter is a browser-based tool for instructors that converts Word documents (.docx) into digitally accessible output. It runs entirely in your browser — no files are uploaded to any server.

Primary use case

Many documents exist only as PDFs with no accessibility structure. Converting a PDF to .docx (via Adobe Acrobat, Smallpdf, ILovePDF, or similar) and running it through this converter is often the only practical path to producing an accessible version of an existing document. This tool is designed to make that workflow as fast as possible.

What this tool checks and fixes

• Document language — sets lang="en-US" in the document XML (automatic)
• Heading structure — detects manually-formatted headings using font size, bold/caps/italic formatting, and text length as signals; proposes Title/H1/H2/H3 styles for review. Boundary paragraphs (References, Acknowledgements, etc.) are also proposed as headings. Already correctly-styled headings are shown in a "Show all" view.
• List formatting — detects manual bullet/number characters and converts to proper Word list styles (automatic, opt-out per item). Reference-section paragraphs are excluded.
• Hyperlinks — three categories:
  • Main text external — bare URLs or generic text; needs descriptive replacement
  • Main text internal — citation numbers [1][2] linking to the reference list; not an accessibility problem, shown collapsed for information only
  • Reference section — links after the References/back-matter heading; collapsed by default, sub-classified as probably-essential (has "accessed/retrieved" language), probably-deletable (DOI/PubMed), or review
• Plain-text URLs — detects URLs that appear as plain text rather than live hyperlinks; offers "Make live" to convert them to clickable links with optional descriptive text
• Images — detects images without alt text; shows thumbnails with AI-prompt and copy-image buttons; separate Description and Title fields; strip per image or strip all
• Tables — detects tables, warns about missing header rows and all-rows-as-headers; shows tab-separated text preview per table; checkbox to auto-insert text rendering after the table in the output DOCX (using a TableTextRendering style)

PowerPoint

A separate Slides tab applies a deliberately minimal fix to .pptx files: white backgrounds with black text, and alt text derived from slide titles for images that have none. It is intended for presentations kept online for reference rather than for active study, as a way of improving the accessibility of material that would otherwise simply be left as it is. It reports, but cannot fix, missing slide titles and text overlaying photographs, and does not address reading order.

Limitations

• Heading detection is heuristic — imperfect, especially in PDF-converted documents. Always review before accepting. Rejected headings are marked with a NotAHeading style so they are not re-flagged on subsequent passes.
• Alt text is not generated automatically — write descriptions yourself; use the AI prompt buttons to get help from an AI assistant.
• Table header markup has known Word 365 bugs — see the Help tab for workarounds including table-to-text and table-to-image approaches.
• Reference section detection — uses text-matching (style-agnostic) and sequential numbering patterns; may not detect all boundaries in unusual citation formats.
• Use iteratively — apply fixes, download, re-upload to verify. This tool does not substitute for your institution's official accessibility checker.

Why these things matter

Headings applied as bold text rather than Word heading styles, images without alt text, plain-text URLs, and tables without header row markup are among the most common reasons documents fail digital accessibility assessments. Screen readers depend on these structural cues to navigate and interpret documents. Remember: there is a difference between digitally accessible, aesthetically acceptable, and perfect — achieving two out of three is often sufficient!

Output formats

• Accessible DOCX — Word document with corrected styles, suitable for further editing or conversion to PDF (use File → Export → Create PDF/XPS → Options → Document structure tags for accessibility)
• Accessible HTML — clean readable HTML suitable for embedding in course pages or sharing as a standalone reading

Privacy

All processing happens in your browser using JavaScript. Your document is never sent to any server.

Citation

Abedon, S.T. Text Doc Digital Accessibility Converter. docaccess.biologyaspoetry.org. DOI: 10.5281/zenodo.20095643