Blog · Reading strategies

Reading data tables online with dyslexia - why grids are hard and what helps

You're comparing mobile phone tariffs. Or reading a public transit schedule. Or trying to make sense of a government statistics page. The page has a table - rows, columns, maybe a dozen numbers across six or eight columns - and you have to track a specific value across to its label without losing the row. For many dyslexic readers, that task is disproportionately exhausting. Not a personal failing; a genuine design and perception mismatch.

This article explains why data tables are one of the hardest formats for dyslexic readers, what makes a well-designed table easier to navigate, and - because most tables online are not well designed - what you can do in your browser right now to make them more manageable.

Why tables are harder than paragraphs

Reading a paragraph is largely linear. Your eye moves left to right, drops to the next line, and continues. It's a single-axis scan. Reading a table requires simultaneous tracking on two axes: you must hold a row in mind as your eye moves right across several columns, and you must return your eye to the start of that same row (not the one above or below) if you need to re-read a cell. That two-axis demand is not a small addition to reading load - for dyslexic readers who already expend extra cognitive effort on the decoding step, it can tip a manageable task into an overwhelming one.

Several specific mechanisms make this worse:

Row slippage. When you look away from the left edge of a row and then look back, the eye has to re-acquire the correct row. For most readers this is automatic. For dyslexic readers who already spend more cognitive resources tracking text position - a well-documented feature of how dyslexia affects eye tracking - the cost of re-acquiring the row is much higher, and errors are more common. You end up reading the next row down, or the one above, and not noticing.

Uniform column typography. In a paragraph, words vary in length, rhythm, and visual texture. A table cell is often a number, a short label, or an abbreviation - all roughly the same visual weight, sitting in equally sized boxes. That uniformity removes the natural irregularity of text that the eye uses as a landmark. Every cell looks like every other cell, and the risk of column drift (reading down instead of across) rises.

Working memory demands. Following a row to its right-most value means holding the row's header in memory while you scan across. Working memory difficulties are common in dyslexia and are compounded by any effort spent on decoding the individual cells. By the time you reach the fifth column, the row label - three or four columns to the left - may have dropped out of working memory entirely. You look back to refresh it, lose your column position, and have to restart the scan.

Headers that scroll away. On long tables, the column headers disappear above the viewport when you scroll down. Now you're reading a grid of numbers with no visible labels, relying entirely on memory of which column is which. Most readers find this annoying; dyslexic readers find it actively unusable.

What makes a table easier to read - and why most online tables fail

Well-designed tables use striped rows (alternating background colours), sticky headers, generous column spacing, and left-aligned text inside cells. These four features are not decoration - each one directly addresses one of the tracking problems described above. Striped rows give the eye a colour-coded lane to follow. Sticky headers mean the column context is always visible. Generous spacing gives the eye room to distinguish adjacent cells. Left-aligned numbers in cells let the eye rest at a consistent starting point in each column.

Table with striped rows - easier to track horizontally
PlanDataCallsPrice/moContract
Basic5 GBUnlimited£1212 months
Standard20 GBUnlimited£1824 months
Plus50 GBUnlimited£2412 months
UnlimitedNo capUnlimited£3530 days
Same data without stripes or borders - much harder to track
PlanDataCallsPrice/moContract
Basic5 GBUnlimited£1212 months
Standard20 GBUnlimited£1824 months
Plus50 GBUnlimited£2412 months
UnlimitedNo capUnlimited£3530 days

Most online tables - government data tables, comparison sites, Wikipedia tables, financial disclosure tables - are not designed to these standards. The page's CSS may strip borders, use hairline dividers that disappear on your screen's contrast level, or set the table in a small serif font at 13px that sits at the edge of legibility. The table was designed for sighted users with no particular reading difficulty and rendered on the screen resolution of the developer's MacBook. It may not look like that on yours.

What you can actually change in the browser

1. Apply a dyslexia-friendly font

Fonts with clear, differentiated letterforms help inside tables for the same reason they help in paragraphs. In a table, every cell is already competing for visual attention with adjacent cells, and letterforms that look similar (1 vs l vs I, 0 vs O, 8 vs B) create an additional layer of ambiguity. A dyslexia-friendly font like Atkinson Hyperlegible - which was specifically designed to distinguish similar characters - or Lexend (which increases glyph width to reduce crowding) can meaningfully reduce cell-level decoding errors.

LexiFont applies your chosen font to all elements on the page, including table cells, with one click. The font override applies at the browser level, so it works on any website you visit - the comparison table on a telco's site, the timetable on your rail operator's site, a government statistics page - without any special setup per site.

2. Increase the base font size

Most browsers let you increase the default font size independently of the zoom level. In Chrome: Settings - Appearance - Font size. Changing this from Medium (16px) to Large (20px) makes table cells physically bigger and increases the visual separation between adjacent rows, which partially compensates for missing row stripes. This is worth trying even if you don't change anything else - font size for dyslexic adults is often underestimated as an intervention.

3. Zoom into the specific table

Ctrl+Plus (or Cmd+Plus on Mac) zooms the entire page. For a table that is otherwise unreadable at the page's default size, zooming in so that only two or three columns are visible at once reduces the horizontal tracking demand. You are now reading a narrower slice of the table and scrolling horizontally to reach the remaining columns. This is slower but more accurate than trying to track across all eight columns at once.

4. Use a physical reading guide

If you use a ruler or piece of paper under lines of text to prevent row slippage, the same technique works on screen. Move the physical edge along the screen surface to mark the row you're reading. This sounds low-tech because it is - but it works for the same reason a browser reading ruler works: it gives the eye an external anchor that removes the working memory cost of tracking your position.

5. Hover highlighting

Some tables respond to hovering by highlighting the row under the cursor. If you hover over a cell, the entire row lights up. This is the most useful single accessibility feature a table can have and it's trivially easy to add in CSS - but many sites don't bother. If a table has hover highlighting, use it deliberately: move the cursor to the row before you start reading across, then keep it there as a position anchor.

6. Copy the table into a spreadsheet

If a table is genuinely complex - many rows, many columns, values you need to compare non-sequentially - copying and pasting it into a spreadsheet program is not a workaround, it is the right tool for the job. Most HTML tables paste cleanly into Google Sheets or Microsoft Excel with Ctrl+C / Ctrl+V. Once the data is in a spreadsheet, you can freeze the top row so headers stay visible, sort by any column, add your own conditional formatting to highlight rows of interest, and delete columns you don't care about. This turns an unnavigable 12-column web table into a filtered, formatted table showing only the three rows and four columns you actually need.

Reading spreadsheets with dyslexia has its own challenges, and we've covered the setup in detail in our guide to reading spreadsheets with dyslexia. But a dyslexia-formatted spreadsheet is almost always easier to navigate than a poorly designed HTML table.

When the font change alone makes the biggest difference

Font matters most in tables where the cells contain abbreviations, codes, or mixed alphanumeric strings - the kind of values where a single misread character changes the meaning. NHS prescription codes, part numbers, IBAN account details, flight numbers, train platform codes - all of these are dense, arbitrary-looking strings where letter shape ambiguity causes real errors.

In these cases, the default font's treatment of 1/l/I, 0/O, 5/S, and 8/B is genuinely consequential. Atkinson Hyperlegible handles all of these with explicitly differentiated forms - it was designed for visually impaired readers where exactly this kind of ambiguity causes practical harm. Switching to it via LexiFont before you read a table of codes, numbers, or reference strings is one of the highest-value uses of a font override tool.

For more on how font clarity interacts with reading accuracy, the guide on contrast sensitivity and dyslexia covers the visual science behind why some letterforms cause more confusion than others.

The colour dimension

Some dyslexic readers - particularly those who also experience visual stress or Irlen-like symptoms - find that grid lines and the box-structure of a table create a distracting "glare" or visual noise on top of the content. If that's your experience, switching to a coloured background (cream, light grey, or pale yellow) through your browser's accessibility settings or a browser extension can reduce the visual noise without affecting the table's structure. Dark mode is worth testing too - some readers find the inverted contrast of white text on dark background significantly reduces table glare - though responses are individual and not universal.

A quick diagnostic: if you find tables specifically hard - not just long text generally - try placing your finger on the screen at the left edge of the row you're reading. If this makes the table manageable, the problem is row tracking, and a reading ruler or hover row highlight is the right fix. If it doesn't help much, the problem is more likely at the cell level (font clarity, size, or contrast), and a font change will do more work.

When the table is an image, not HTML

Some older government documents, PDF reports converted to HTML, and scanned pages display tables as images rather than real HTML elements. In this case, none of the font or zoom tricks help with the cells themselves - the content is a pixel image and the browser cannot reformat it. Your options are:

First, try increasing zoom anyway - a bigger image is at least physically easier to scan, even if the font doesn't change. Second, look for a "download as PDF" or "view original" option - a real PDF may have selectable text that you can copy out. Third, if the document is important, a screen reader with OCR capability (or an AI tool that accepts images) can extract the table content into plain text that you can then reformat. This is the route of last resort but it works when the alternative is struggling through a 12-column image table at 100% zoom.

Making it routine

Most of the fixes above work best when they're set up in advance rather than reached for in frustration. If you read tables regularly - for work, for comparing services, for research - the fifteen-minute investment in setting up LexiFont Pro with your preferred dyslexia font, adjusting your browser's base font size, and learning the keyboard shortcut for zoom pays back quickly. You don't want to be debugging your reading setup while you're also trying to work out which train platform you're supposed to be on.

The workflow for reading long articles with dyslexia covers the same principle of front-loading your setup. The same logic applies to any reading task that involves a demanding layout - set the environment before you start reading, not mid-way through.

Further reading