Lab 03: IBM Equal Access — Comprehensive Policy Scanning

Duration	30 minutes
Level	Intermediate
Prerequisites	Lab 01

Learning Objectives

By the end of this lab, you will be able to:

• Explain how IBM Equal Access differs from axe-core in rule coverage and approach
• Run an IBM Equal Access scan on a demo app using the scanner
• Compare IBM findings with axe-core findings for the same page
• Review a combined report that deduplicates findings across engines

Exercises

Exercise 3.1: Understand IBM Equal Access

The scanner supports two accessibility engines. You will review how IBM Equal Access complements axe-core.

1. Review the engine comparison:

   Aspect	axe-core	IBM Equal Access
   Maintainer	Deque Systems	IBM
   Rule count	~90 rules	~400+ rules
   Focus	WCAG conformance testing	Policy-based assessment (WCAG + IBM requirements)
   Result types	Violation / Pass / Incomplete	Violation / Need Review / Recommendation
   Strengths	Fast, industry standard, low false-positive rate	Broader rule coverage, policy customization, government compliance

2. Key differences to note:

   • IBM Equal Access uses the ACE (Accessibility Conformance Engine) which includes rules beyond WCAG, such as IBM-specific requirements.
   • IBM categorizes findings as Violation (definite failure), Need Review (potential issue requiring human judgment), and Recommendation (best practice suggestion).
   • The scanner’s accessibility-checker npm package provides the IBM engine integration.

[!NOTE] Both engines are complementary. axe-core catches common violations with high confidence, while IBM Equal Access provides broader coverage and catches issues axe-core may miss.

Exercise 3.2: Run IBM Scan on Demo App 002

You will scan a demo app using the IBM Equal Access engine.

1. Ensure demo app 002 is running at http://localhost:8002.

2. Open the scanner web UI at http://localhost:3000.

3. Enter the demo app 002 URL and select the scan options. If the scanner supports engine selection, choose IBM Equal Access or Combined mode.

4. Review the results. IBM Equal Access typically finds additional issues that axe-core does not flag, such as:
   • Elements that need review for keyboard operability
   • Advisory recommendations for ARIA usage
   • Policy-specific checks for form interactions

   

5. Click on a specific IBM finding to view its details. Note the difference in rule identifiers — IBM rules use identifiers like WCAG20_Html_HasLang while axe-core uses html-has-lang.

   

Exercise 3.3: Compare IBM and axe-core Findings

You will compare the findings from both engines on the same page.

1. Review the side-by-side comparison for demo app 002:

   Category	axe-core	IBM Equal Access
   Total findings	~20–30 violations	~40–60 violations + need-review
   Language check	html-has-lang	WCAG20_Html_HasLang
   Image alt text	image-alt	WCAG20_Img_HasAlt
   Color contrast	color-contrast	IBMA_Color_Contrast_WCAG2AA
   Form labels	label	WCAG20_Input_ExplicitLabel
   Unique to engine	Keyboard trap detection	Need-review items, recommendation items

   

2. Note that some violations are detected by both engines under different rule names. The scanner normalizes and deduplicates these in the combined report.

3. Pay attention to findings that appear only in IBM Equal Access. These are often related to:

   • ARIA attribute validation
   • Widget role definitions
   • Proper use of tabindex
   • Reading order and focus management

Exercise 3.4: Review Combined Report

The scanner can merge findings from both engines into a single deduplicated report.

1. Run a combined scan (both engines) on demo app 002 via the CLI:

   npx ts-node src/cli/commands/scan.ts --url http://localhost:8002 --format json --output results/demo-002-combined.json
   

2. Open results/demo-002-combined.json and review the structure. The combined report:
   • Lists each unique violation once, even if both engines detected it
   • Indicates which engine(s) flagged each finding
   • Preserves the highest impact level when engines disagree

   

3. Review how deduplication works:

   

   • Rules are matched by their target WCAG success criterion
   • When both engines find the same violation on the same element, a single entry is created
   • Engine-specific metadata is preserved for traceability

[!TIP] Using both engines together provides the most comprehensive coverage. axe-core’s low false-positive rate combined with IBM’s broader rule set catches violations that either engine alone would miss.

Verification Checkpoint

Before proceeding, verify:

• Can explain the differences between axe-core and IBM Equal Access
• Ran an IBM scan on demo app 002 and reviewed the findings
• Identified at least 2 findings that IBM catches but axe-core does not
• Reviewed a combined report showing deduplicated findings from both engines

Next Steps

Proceed to Lab 04: Custom Playwright Checks — Manual Inspection.