Evaluation of ergonomic errors in the interaction with computer mice
A.1 Purpose and aim
III. LOGITECH MARBLE MOUSE USB
Figur A3 LOGITECH MARBLE MOUSE USB
This mouse was sold with the following Swedish text: “Du sparar plats och minskar belastningen på hand och handled. Den rundade formen gör pekdonet väl lämpat för långa arbetspass.” http://www.dustin.se/pd_5010169230.aspx (Dustin, 2011)
Figur A4 3M ERGONOMIC OPTICAL MOUSE SMALL and LARGE
This mouse was sold with the following Swedish text: “Ergonomisk optisk mus med vertikalt grepp. Kliniskt utprovad för att reducera muskelvärk och obehag jämfört med en traditionell mus. Motverkar obehag som på sikt kan leda till musarm.” (Dustin, 2011). Here, two sizes were used according to the manufacturer's recommendation.
A.3 Method
First the intended use of a computer mouse (Figure A1) was mapped with HTA (Hierarchical Task Analysis) (Stanton, 2006). A task analysis is needed as a base for the PEEA evaluation and also to structure the results from a usability study.
Grab and hold Click and scroll Moving and controlling Use computer mouse
Figure A1 HTA of mouse usage
When the general HTA was set, an analytical evaluation of physical ergonomic errors in the interaction between the user and product was made PEEA. The evaluation resulted in a list of possible ergonomic errors associated with the various stages of use for each of the four
computer mice selected to be tested. The evaluation was performed by two analysts (LOB and ALO).
The next step was to perform usability tests with the four computer mice. 16 people were recruited from a department at Chalmers University. The test subjects did not have deeper knowledge of ergonomics, neither suffered they from muscular-skeletal problems in the hand and arm region. These criteria were set because it was judged to affect the outcome of the results if they had deeper knowledge of ergonomic issues than an average office worker. There were five women and eleven men included, with a mean age of 34.5 years, reflecting age and gender composition of the department staff.
The 16 subjects used three of the mice to solve given tasks. Which mice and their order in the test session varied to avoid systematic errors. The decision that each test subject should only test three of the four mice was time usage and that the test subjects should not get too bored, exhausted or uninterested by performing the same tasks too many times.
The tasks which should be made with each mouse were: 1. Open the file mustest.docx on the desktop 2. Change to the current date in the header
3. Type your name under the heading on the other side 4. Cut the last paragraph on the last page
5. Paste the paragraph under your name on the other side
6. Save the file under a new name "mustest A.docx" on your desktop 7. Close Word
8. Open Chalmers webmail and login (link on the desktop) 9. Open a new email
10. Enter the address "lars-ola.bligard @ chalmers.se" and the subject "Mustest". 11. Enter the text "Hello! Greetings <your name> "in the email.
12. Attach the file "mustest.docx"
13. Turn on the high priority on the email 14. Send the email
15. Close the browser
Each test session was filmed with focus on hands and arms, in order to document how the test subjects used the different mice when they performed the tasks. The films were later analysed to identify use errors that the subjects did when handling the mice, i.e. if they handled the mouse in an incorrect ergonomic way compared to how it was intended to be used. The analysis of the film was conducted by one analyst (MR), who did not participate in the
analytical evaluation with PEEA. Finally, the ergonomic errors found from the PEEA analysis were compared with the errors found in the usability tests. The comparison is reported in the results.
A.4 Results
The tables A1-A4 show all the ergonomic errors found by PEEA and in the usability tests for each computer mouse tested. On the left hand side in the tables the tasks described in the HTA are listed. Below each task there is a list of all use errors found with PEEA or in the usability tests. On the right hand side the numbers show how many test subjects who performed each use error. When the number is zero (0), no test subject performed that error during the usability session. The errors that occurred during the usability tests but were not found by PEEA are marked with grey.
Table A1 Ergonomic errors with AHAA VERTICALMOUSE
Grab and hold No.
Grab obliquely 15
Focus too far back (grasping for a bit) 5
Grab with left hand 0
Grab too low 10
Holding the hand above the mouse (fingers on the keys) 4
Keeping the fingers over the mouse (hand on the mouse) 0
Keeping the wrist above the table 1
Click and scroll
Index finger on the right button 0
Ring finger on the mouse buttons 0
Index finger is placed to the left of the left button 0
Place the fingers too far down 3
Scroll with the ring finger 0
Working with bent fingers 0
Keep fingers above the keys between clicks 0
Moving and controlling
Working with a bent wrist. 7
Control with the wrist instead of the entire forearm 12
Not having wrist and forearm on the table 2
Not placing the elbow on the table (if the elbow is over the table) 0
Table A1 shows that PEEA had produced 18 possible ergonomic errors. Of these, 9 occurred in the usability tests. Four of the errors occurred more frequently (> 50% of the test subjects performed the error). No other ergonomic errors than those predicted with PEEA occurred during the tests with AHAA Vertical mouse.
Table A2 Ergonomic errors with EVOLUENT VERTICAL MOUSE 4 ERGONOMIC USB RIGHT
Grab and hold No.
Grab obliquely 5
Focus too far back (grasping for a bit) 4
Grab with left hand 0
Grab too low 1
Grab too high 3
Grab too far forward 1
Holding the thumb beneath thumb grip 1
Holding the hand above the mouse (fingers on the keys) 5
Keeping the fingers over the mouse (hand on the mouse) 2
Keeping the wrist above the table 1
Click and scroll
Index finger used for the right or the middle button 0
Index finger is placed to the left of the left button 3
Place the fingers too low 0
Scroll with ring finger 0
Working with bent fingers 3
Keep fingers above the keys between clicks 0
Clicking too high on the left button, almost on top 2
Clicking on the middle button with the ring finger 1
Clicking on the left button with middle finger 2
Moving and controlling
Working with a bent wrist. 2
Controlling with the wrist instead of the entire forearm 12
Not having wrist and forearm on the table 1
Not placing the elbow on the table (if the elbow is over the table) 0
Table A2 shows that PEEA had produced 17 potential ergonomic errors. Of these, 11 occurred in the usability tests. One of the errors occurred frequently (> 50% of the test subjects performed the error). For this mouse there were 7 other ergonomic errors found during the usability tests that PEEA had not predicted. These were all related to the clicking function of the mouse.
Table A3 Ergonomic errors with LOGITECH MARBLE MOUSE USB
Grab and hold #
Grab obliquely (across) 5
Grab too far back 0
Holding hand above the mouse (fingers on the keys) 1
Keeping fingers over the mouse (hand on the mouse) 12
Keeping the wrist above the table 2
Keeping far ahead, above the ball 1
Click and scroll
Clicking with the left index finger on the button instead of the thumb 0
Clicking with index finger on right key 0
Placing fingers near the bottom left 0
Placing fingers too far down to the right 2
Controlling the ball with ring finger or little finger 1
Working with bent fingers 3
Keep fingers above the keys between clicks 12
Moving and controlling
The user attempts to move the whole mouse 0
Using too many fingers on the ball 0
Working with a bent wrist. 2
Not having forearm on the table 1
Not placing elbow on the table (if the elbow is over the table) 0
Table A3 shows that PEEA had produced 18 possible ergonomic errors. Of these, 11 occurred in the usability tests. Two of the errors occurred more frequently (> 50% of the test subjects performed the error). No other ergonomic errors than those predicted by PEEA occurred during the usability tests.
Table A4 Ergonomic errors with 3M ERGONOMIC OPTICAL MOUSE SMALL and LARGE
Grab and hold #
Grab obliquely 0
Placing fingers on the side instead of around 1
Keeping thumb in the air 2
Working with a bent thumb 1
Keeping the whole hand in the air (in anticipation of continuing to use the joystick)
1
Grabbing with left hand 0
Grab far down on the bottom plate 0
Putting hand on top of the joystick 0
Grabbing far down on the joystick 2
Grabbing near the top of the joystick 0
Click and scroll
Using thumb to press the side button instead of the other fingers 1
Using other fingers than thumb to press the thumb-button 0
Moving and controlling
Try to bend or twist the grip 0
Working with a bent wrist. 1
Controlling with the wrist instead of the entire forearm 8
Not placing the elbow on the table (if the elbow is over the table) 2
Table A4 shows that PEEA had predicted13 potential ergonomic errors. Of these, six occurred in the usability tests. Only one of the errors occurred frequently (> 50% of the test subjects performed the error). 3 other ergonomic errors occurred during the usability tests, most related to the handling of the thumb.
A.5 Analysis
To investigate how well the PEEA method worked, the use errors found with the different evaluation methods for the four mice were compared. The errors were compared on the basis of: (1) Number of different ergonomic errors performed in the usability tests, (2) Number of errors found in the usability tests predicted by PEEA, (3) Number of errors predicted by PEEA that not occurred during the usability tests, and (4) Number of usability test errors that not was predicted by PEEA. In Table A5 the use error comparison is shown.
Table A5 Comparison of use errors
All errors for all mice Mouse
1 Mouse 2 Mouse 3 Mouse 4 In total
1. Errors in the usability tests 9 18 11 9 47
2. Usability test errors predicted by PEEA 9 11 11 6 37
3. Predicted by the PEEA not occurred 9 6 7 7 29
4. Incurred which is not predicted by the PEAA
0 7 0 3 10
Common errors (> 50% of test subjects) Number of different errors in the
experiments
4 1 2 1 8
Of those predicted by PEEA 4 1 2 1 8
Of those is not predicted by PEAA 0 0 0 0 0
The comparison showed that more than half (50%, 65%, 58%, 46%) of the predicted errors by PEEA occurred in a real life situation. For two of the mice, PEEA identified all errors that occurred. For the other mice, 61% and 67% respectively of the errors were predicted. For the most common errors, where more than 50% of the test subjects did this error, the PEEA method predicted them all.
For two of the mice, no other ergonomic errors than those PEEA predicted occurred during the usability tests (Tables A1 and A3). For the other two mice (Tables A2 and A4), 7 and 3 errors respectively occurred during the usability test, these were not found with PEEA in the analytical evaluation. For these two mice, the analysts who used PEEA did not consider all the different ways that wrong grips and clicking could be made, nor how the thumb could be handled in a wrong position. The analysts did not perform the tasks that were involved in the usability tests during their evaluation of presumptive use errors, and that can be one of the causes that these errors were not expected.
Table A6 In total: Use errors found by PEEA and in usability test
Errors Errors found by
PEEA
Errors not found by PEEA
In total:
Errors found in usability tests 37 10 47
Errors not found in usability tests 29 - 29
In total: 66 10 76
An interesting comparison is between the total number of errors predicted by PEEA for all mice and the total number of errors that occurred during the usability tests for all mice (Table A6). Of the 47 errors that occurred during the usability tests PEEA had predicted 37, i.e. 79%. Of the 66 errors predicted by PEEA, 37 occurred during the tests, i.e. 56%. This is a good result because it is more important that PEEA finds the errors that actually occur, then that the
errors predicted by PEEA are real use errors. A remaining question is how many of the other use errors that PEEA predicted would occur if more test subjects had been involved in the tests? Also interesting to ask is if and how many ergonomic errors neither PEEA nor the usability test found. However, the most frequent and usual errors might have been found with a combined analysis of PEEA and usability test.
A.6 Conclusion
The comparison has shown that PEEA worked well at predicting the ergonomic errors that occurred when handling ergonomic computer mice, especially for the most frequent errors. PEEA has predicted more errors than occurred, but that lies in the nature of the method to find as many errors as possible. By this approach, the purpose is to minimise the number of severe errors that can occur in an interaction between user and artefact.
Another aspect, which was not the purpose of this specific study, is that as many as 76 presumptive ergonomic errors can occur when users work with computer mice sold as being ergonomic. This highlights the importance of informing users how to handle ergonomic products in a correct way and also develop ergonomic products with a high degree of guessability for correct usage. The best design idea is that a product only can be used in one possible way, the correct way, bit this is nearly impossible to achieve.
A.7 References
DUSTIN. 2011. Dustin Webpage [Online]. Stockholm. Available: www.dustin.se [Accessed 2011-03-12].
STANTON, N. A. 2006. Hierarchical task analysis: Developments, applications, and extensions. Applied Ergonomics, 37, 55-79.