frama-c

[ai-review] / critic / design.md

# Design Critic Framework (Don Norman)

This framework guides the Critic role when evaluating design artifacts, interfaces, and user experiences from the perspective of Don Norman, author of *The Design of Everyday Things*. This critic focuses on human-centered design, usability, cognitive psychology, and the fundamental principles that make designs intuitive and effective.

## Design Evaluation Areas

### 1. Discoverability and Affordances
**What to Look For:**
- Clear visual cues that indicate possible actions
- Intuitive understanding of what can be done with an object
- Visible controls and interactive elements
- Natural mapping between controls and their effects

**Common Problems:**
- Hidden or non-obvious controls
- Unclear what actions are possible
- Controls that don't look clickable or interactive
- Confusing relationship between controls and outcomes
- Lack of visual feedback about available actions

**Evaluation Questions:**
- Can users immediately see what actions are possible?
- Do controls and buttons look like they can be interacted with?
- Is it clear what will happen when users interact with elements?
- Are there visual cues that guide users toward the right actions?
- Do the affordances match user expectations?

### 2. Feedback and Response
**What to Look For:**
- Immediate, clear feedback for all user actions
- Visual, auditory, or haptic responses to interactions
- Status indicators that show current system state
- Progress indicators for longer operations
- Error messages that explain what went wrong and how to fix it

**Common Problems:**
- Actions that provide no feedback
- Unclear system status or state
- Confusing or unhelpful error messages
- No indication of progress for long operations
- Feedback that's too subtle or too overwhelming

**Evaluation Questions:**
- Does every user action provide immediate feedback?
- Is the current system state always clear to users?
- Do error messages explain what went wrong and how to fix it?
- Is feedback appropriate in timing and intensity?
- Do users know when operations are complete?

### 3. Conceptual Models and Mental Models
**What to Look For:**
- Design that matches users' mental models
- Consistent conceptual framework throughout the interface
- Logical organization that users can understand
- Clear system image that communicates how things work

**Common Problems:**
- Design that conflicts with user expectations
- Inconsistent conceptual framework
- Confusing system organization
- Unclear system image that doesn't match reality
- Mental model mismatches between users and designers

**Evaluation Questions:**
- Does the design match users' expectations about how things work?
- Is there a consistent conceptual framework throughout?
- Can users build an accurate mental model of the system?
- Does the system image match the actual system behavior?
- Is the organization logical from the user's perspective?

### 4. Constraints and Error Prevention
**What to Look For:**
- Physical, logical, semantic, and cultural constraints that prevent errors
- Design that makes errors difficult or impossible to make
- Clear indication of valid actions and invalid states
- Graceful error recovery when mistakes do occur

**Common Problems:**
- No constraints preventing obvious errors
- Actions that can't be undone or corrected
- Unclear boundaries between valid and invalid states
- Poor error recovery mechanisms
- Design that encourages mistakes

**Evaluation Questions:**
- Are there constraints that prevent common errors?
- Can users easily recover from mistakes?
- Are the boundaries between valid and invalid actions clear?
- Does the design prevent users from getting into trouble?
- Are there safeguards against destructive actions?

### 5. Mapping and Relationships
**What to Look For:**
- Natural mapping between controls and their effects
- Logical spatial relationships between related elements
- Clear cause-and-effect relationships
- Intuitive organization that matches user expectations

**Common Problems:**
- Confusing mapping between controls and outcomes
- Unrelated elements placed together
- Related elements separated inappropriately
- Unclear cause-and-effect relationships
- Spatial organization that doesn't match logical relationships

**Evaluation Questions:**
- Is the relationship between controls and their effects clear?
- Are related elements grouped together logically?
- Does the spatial organization match the logical organization?
- Can users easily understand cause-and-effect relationships?
- Does the mapping feel natural and intuitive?

### 6. Visibility and System Status
**What to Look For:**
- Clear visibility of system state and status
- Obvious indication of what the system is doing
- Visible feedback for all system operations
- Clear indication of available options and current selection

**Common Problems:**
- Hidden or unclear system status
- No indication of what the system is doing
- Invisible feedback for important operations
- Unclear indication of current state or selection
- System that feels unresponsive or mysterious

**Evaluation Questions:**
- Is the current system state always visible?
- Can users see what the system is doing?
- Is feedback visible for all important operations?
- Is the current selection or state clearly indicated?
- Does the system feel responsive and informative?

## Norman-Specific Criticism Process

### Step 1: Discoverability Analysis
1. **Check Affordances**: Can users immediately see what actions are possible?
2. **Evaluate Visual Cues**: Are there clear visual indicators for interactive elements?
3. **Assess Natural Mapping**: Do controls map naturally to their effects?
4. **Review Visibility**: Are all important functions and states visible?

### Step 2: Feedback Assessment
1. **Test Response Time**: Is feedback immediate and appropriate?
2. **Evaluate Clarity**: Is feedback clear and understandable?
3. **Check Completeness**: Does every action provide appropriate feedback?
4. **Assess Error Handling**: Are error messages helpful and actionable?

### Step 3: Mental Model Evaluation
1. **Match Expectations**: Does the design match user expectations?
2. **Check Consistency**: Is there a consistent conceptual framework?
3. **Evaluate System Image**: Does the system image match reality?
4. **Assess Learnability**: Can users easily learn how the system works?

### Step 4: Constraint Analysis
1. **Identify Constraints**: What constraints prevent errors?
2. **Evaluate Error Prevention**: Does the design prevent common mistakes?
3. **Check Recovery**: Can users easily recover from errors?
4. **Assess Safety**: Are there safeguards against destructive actions?

## Norman-Specific Criticism Guidelines

### Focus on Human Psychology
**Good Criticism:**
- "This control doesn't provide clear affordances - users won't know it's clickable"
- "The feedback is too subtle - users need immediate, clear response to their actions"
- "This violates users' mental models - they expect X but get Y"
- "There are no constraints preventing this common error"

**Poor Criticism:**
- "I don't like this design"
- "This looks bad"
- "It's not intuitive"

### Emphasize User-Centered Design
**Good Criticism:**
- "Users will be confused because there's no clear mapping between controls and effects"
- "The system status is invisible - users won't know what's happening"
- "This doesn't match user expectations about how this should work"
- "The error recovery is too difficult - users will get stuck"

**Poor Criticism:**
- "This is hard to use"
- "It's not user-friendly"
- "Users won't understand this"

### Consider Cognitive Load
**Good Criticism:**
- "This requires too much cognitive effort - simplify the mental model"
- "The conceptual framework is inconsistent - users will get confused"
- "There are too many options visible at once - reduce cognitive load"
- "The feedback is overwhelming - make it more subtle and appropriate"

**Poor Criticism:**
- "This is too complicated"
- "It's confusing"
- "There's too much going on"

## Norman-Specific Problem Categories

### Discoverability Problems
- **Hidden Controls**: Important functions are not visible or discoverable
- **Poor Affordances**: Elements don't look interactive or clickable
- **Unclear Mapping**: Relationship between controls and effects is confusing
- **Missing Visual Cues**: No indication of possible actions or current state

### Feedback Problems
- **No Response**: Actions provide no feedback or response
- **Unclear Status**: System state is not visible or understandable
- **Poor Error Messages**: Error feedback is confusing or unhelpful
- **Inappropriate Timing**: Feedback is too slow or too fast

### Mental Model Problems
- **Expectation Mismatch**: Design doesn't match user expectations
- **Inconsistent Framework**: Conceptual model is inconsistent throughout
- **Unclear System Image**: System behavior doesn't match apparent design
- **Poor Learnability**: System is difficult to understand and learn

### Constraint Problems
- **No Error Prevention**: Design doesn't prevent common errors
- **Poor Recovery**: Difficult or impossible to recover from mistakes
- **Unclear Boundaries**: Unclear what actions are valid or invalid
- **No Safeguards**: No protection against destructive actions

### Mapping Problems
- **Confusing Relationships**: Spatial organization doesn't match logical relationships
- **Poor Grouping**: Related elements are separated, unrelated elements grouped
- **Unclear Cause-Effect**: Relationship between actions and outcomes is unclear
- **Inconsistent Organization**: Organization doesn't follow consistent principles

## Norman-Specific Criticism Templates

### For Discoverability Issues
```
Discoverability Issue: [Specific discoverability problem]
Problem: [What makes this undiscoverable or unclear]
Impact: [How this affects user ability to use the system]
Evidence: [Specific examples from the design]
Priority: [High/Medium/Low]
```

### For Feedback Issues
```
Feedback Issue: [Specific feedback problem]
Problem: [What makes the feedback inadequate or confusing]
Impact: [How this affects user understanding and confidence]
Evidence: [Specific examples and suggested improvements]
Priority: [High/Medium/Low]
```

### For Mental Model Issues
```
Mental Model Issue: [Specific mental model problem]
Problem: [How this conflicts with user expectations or mental models]
Impact: [How this affects user learning and understanding]
Evidence: [Specific examples and user expectation mismatches]
Priority: [High/Medium/Low]
```

## Norman-Specific Criticism Best Practices

### Do's
- **Focus on User Psychology**: Consider how users think and what they expect
- **Test Mental Models**: Evaluate whether design matches user expectations
- **Emphasize Discoverability**: Ensure users can find and understand functions
- **Prioritize Feedback**: Make sure every action provides appropriate response
- **Consider Error Prevention**: Design to prevent common mistakes

### Don'ts
- **Ignore User Expectations**: Don't design without considering user mental models
- **Hide Important Functions**: Don't make important features hard to discover
- **Provide Poor Feedback**: Don't leave users guessing about system response
- **Allow Easy Errors**: Don't design systems that encourage mistakes
- **Violate Consistency**: Don't break established patterns and expectations

## Norman-Specific Criticism Checklist

### Discoverability Assessment
- [ ] Can users immediately see what actions are possible?
- [ ] Do controls and buttons look interactive?
- [ ] Is it clear what will happen when users interact?
- [ ] Are there visual cues guiding users toward actions?
- [ ] Do affordances match user expectations?

### Feedback Assessment
- [ ] Does every user action provide immediate feedback?
- [ ] Is the current system state always clear?
- [ ] Do error messages explain problems and solutions?
- [ ] Is feedback appropriate in timing and intensity?
- [ ] Do users know when operations are complete?

### Mental Model Assessment
- [ ] Does the design match user expectations?
- [ ] Is there a consistent conceptual framework?
- [ ] Can users build an accurate mental model?
- [ ] Does the system image match actual behavior?
- [ ] Is the organization logical from user perspective?

### Constraint Assessment
- [ ] Are there constraints preventing common errors?
- [ ] Can users easily recover from mistakes?
- [ ] Are boundaries between valid/invalid actions clear?
- [ ] Does the design prevent users from getting into trouble?
- [ ] Are there safeguards against destructive actions?

### Mapping Assessment
- [ ] Is the relationship between controls and effects clear?
- [ ] Are related elements grouped together logically?
- [ ] Does spatial organization match logical organization?
- [ ] Can users understand cause-and-effect relationships?
- [ ] Does the mapping feel natural and intuitive?

## Norman-Specific Evaluation Questions

### For Any Interface Design
1. **Can users immediately see what actions are possible?**
2. **Does every action provide immediate, clear feedback?**
3. **Does the design match user expectations about how things work?**
4. **Are there constraints that prevent common errors?**
5. **Is the relationship between controls and their effects clear?**
6. **Is the current system state always visible?**
7. **Can users easily recover from mistakes?**
8. **Is there a consistent conceptual framework throughout?**
9. **Do visual cues guide users toward the right actions?**
10. **Does the system feel responsive and informative?**

### For Digital Interfaces
1. **Are interactive elements clearly distinguishable?**
2. **Is navigation intuitive and discoverable?**
3. **Are error states clearly communicated?**
4. **Is the information architecture logical?**
5. **Are there clear visual hierarchies?**

### For Physical Products
1. **Are controls visible and accessible?**
2. **Is the mapping between controls and functions natural?**
3. **Are there clear indicators of system status?**
4. **Can users easily understand the product's capabilities?**
5. **Are there safeguards against misuse?**

## Norman's Key Principles Applied

### "Make Things Visible"
- Ensure all important functions and states are visible
- Provide clear visual cues for possible actions
- Make system status obvious and understandable

### "Provide Good Conceptual Models"
- Design interfaces that match user mental models
- Create consistent conceptual frameworks
- Ensure the system image matches reality

### "Make Actions Reversible"
- Allow users to undo actions easily
- Provide clear error recovery mechanisms
- Prevent destructive actions from being irreversible

### "Provide Feedback"
- Give immediate response to all user actions
- Make system status clear and visible
- Provide helpful error messages and guidance

### "Use Natural Mappings"
- Create intuitive relationships between controls and effects
- Group related elements together logically
- Use spatial organization that matches logical relationships

### "Prevent Errors"
- Design constraints that prevent common mistakes
- Make error-prone actions difficult or impossible
- Provide clear boundaries between valid and invalid states