Thumb Reach Distances and Envelopes for Handheld Devices

There are a growing number of handheld devices that involve thumb activated interfaces. Edward W. Otten (Research and Engineering Development, LLC), Keith S. Karn and Kelley S. Parsons (Eastman Kodak) investigated thumb reach envelopes and produced data that is useful to handheld device designers, as well as a methodology they suggest will help others collect thumb reach envelope data for their specific applications.

Even if you are not involved in handheld device design, you likely have first hand experience with them in your own life and may be called upon to assist coworkers or clients with the ergonomics related purchasing or use criteria. This study may be helpful in that effort, and furthermore, is a good case study for how an effective human-centered consumer oriented design project can be carried out.

Introduction

In a collaborative effort between industrial design and ergonomics, Otten et al performed this study as part of a handheld pocket sized camera design project. The design requirements included:

General Device Specifications

• pocket sized;
• held primarily in the palm of one hand, held in a vertical position with the lon-axis perpendicular to the fingers;
• rectangular in shape; and
• use of hardware buttons rather than touch screen (software) activation.

General Task Requirements

• hold the device in the palm with the fingers; and
• accommodate a variety of users performing a "considerable amount of interaction" with the thumb.

Target User Population

• North American;
• adolescent females; and
• adult males and females that are parents of young children. 

Otten et al began by reviewing anthropometry hand, finger and thumb data and studies. They found 50 sources in the published literature, but all but one focused on adults, and all but a few presented only static or structural data with measurements between anatomical landmarks in "unnatural postures." They also reviewed dynamic, functional anthropometry capturing larger multi-body-segment reach envelopes, but found most of them to be more applicable to larger-scale designs such as workstations or cockpits. They only identified a few studies that included dynamic/functional thumb reach data, and say:

The relatively limited data related to thumb reach surprised us, given the indisputable prevalence of handheld electronic devices that are primarily operated with the thumb. The increased use of such devices has resulted in ergonomic problems … including tendonitis from overuse … Despite these issues, we found very few standards or guidelines for the design of devices that rely on thumb input …

Interested readers are directed to the complete research article, cited and linked to below, for a brief review of the few thumb-related studies they did find. The previous research findings did identify some performance information Otten et al found useful:

• participants moved their thumbs faster, and also rated the task easier, when moving between targets on a northeast-to-southwest line; and
• participants moved their thumbs slower, and also rated the task more difficult, when moving between targets on a northwest-southeast line.

However, none of the studies they reviewed provided them with the data they needed for their particular design criteria, so they embarked on their own study using a version of the "Mr. Potato Head" control location method and touch screen method to gather 2-dimensional thumb reach envelopes.

Methods

Participants

• 35 adolescent females (14-19 years old);
• 35 young mothers and 20 young fathers (25-43 years old with at least one child living at home); and
• 30 of the 35 adolescent females, 34 of the 35 young mothers, and 19 of the 20 young fathers were right-handed.

Mr. Potato Head (for control locations)

The researchers constructed a model representing the size and shape of the camera and used magnetic representations of controls, such as record, navigate through, delete and control pictures/videos. Participants were allowed to select among various types and sizes of controls and placed either three or four magnetic controls on the model. Participants were encouraged to 'design' their own cameras, for themselves, and not consider the preferences they might expect for others. The results of this assisted the team in identifying control preferences and locations, but this data is not included in the published article, presumably for confidentiality purposes.

Thumb and Reach Envelope

The experimenters measured hand anthropometric dimensions of the dominant hand using a caliper and a ruler.

An iPod Touch (Apple, Inc.) running a painting application (Sketchbook Mobile, from Autodesk, Inc) was encased in a custom representation of the approximate camera and used to record 2-dimensional reach envelopes. Participants were instructed to hold the device 'securely but comfortably' with their dominant hand, then swipe the thumb across the screen, without changing hand position in an effort to reach further. The shaded area created was then saved and analyzed later to gather reach distance data. Maximum single point reach distances were calculated, as well as a "heat map" type grid that captured 2-dimensional reach coverage.

Results

Key findings include:

Anthropometry

• hand length for young fathers (Mean = 192 mm, SD = 7.1 mm) was significantly larger than for young mothers;
• hand length for young mothers (Mean = 174 mm, SD = 7.1 mm) was significantly larger than for adolescent females (Mean = 167 mm, SD = 12 mm);
• thumb length for young fathers (Mean = 72.8 mm, SD = 17 mm) was significantly larger than young mothers (Mean = 60.6 mm, SD = 5.3 mm) and adolescent females (Mean = 56.1 mm, SD = 6.4 mm), but there was no significant difference between young mother and adolescent female thumb lengths; and
• for control placement, there was no significant differences in preferences between young fathers, young mothers, or adolescent females.

Thumb Reach

• maximum thumb reach distance was generally greater for people with longer thumbs (this was statistically significant, though the correlation was weak);
• young mothers had a significantly smaller thumb reach (Mean = 51.8 mm, SD = 7.1 mm) than young fathers or adolescent females;
• there was no significant difference between maximum thumb reach for young fathers (Mean = 61.4 mm, SD = 9.7 mm) and adolescent females (Mean = 58.9 mm, SD = 11.7 mm);
• there was no single area that all participants could reach comfortably;
• 90% of the participants were able to reach an area between 45.6 mm and 49.4 mm from the bottom and between 25.4 mm and 35.5 mm from the right;
• 70% of the participants were able to reach an area between 34.2 mm and 56.1 mm from the bottom and 15.2 mm and 45.6 mm from the right;
• adolescent females had the largest 2-dimensional reach envelope, and as a whole produced a circular heat map pattern;
• as a group, young fathers produced an elliptical reach envelope;
• as a group, young mothers produced an reach envelope somewhere between a circle and an ellipse, and also showed the smallest reach envelope, with particular difficulty reaching the upper portion of the heat map grid;
• none of the young mothers were able to reach the upper left corner; and
• overall, the long-axis of the reach envelope ellipse was in a northeast-southwest direction.

Preferred Control Locations Compared With Reach Envelopes

• all three demographic groups generally preferred vertical control locations within the area suggested by the heat maps; 
• adolescent females preferred the arrangement slightly higher than the heat map 'sweet spot' suggested;
• young fathers placed controls in the smallest cluster;
• adolescent females preferred the greatest control distance separations;
• young mothers seemed to prefer the controls closely aligned to the central vertical axis.

Interested readers are directed to the full article, cited below, for complete detailed results.

Reviewer Comments

Just as Otten et al expressed, I found it interesting that there is so little publicly available data relating to thumb control interfaces, given that such devices are prevalent in the marketplace. Perhaps companies that produce thumb controlled devices are doing their own internal research and treating it as proprietary trade secret. Or, as is so often the case with technology driven marketplace competition, designers are forging ahead based on what "feels right" to themselves, pushing devices to market that are "close enough," yet fail to perform well in the hands of their customers. In the case of handheld devices, that failure could be in control accuracy or in health concerns among users. Anecdotal experience suggests control errors are common (e.g., lots of "typos" and "fat finger" errors), and that there may be emerging problems hand/thumb/finger discomfort and injury, especially among younger users that have embraced hand-held technologies.

I also found this to be a good example of what I call "informed design." In my view, informed design means the design team recognized and accommodated for the specific target user populations capabilities and preferences, understood and mocked-up the expected user tasks, and designed the product accordingly. A human-centered design approach is the first step in informed design.

I also found it interesting that the three demographic groups displayed as much difference as they did, especially in terms of functional reach distance and 2-dimensional reach envelopes. For example, the adolescent females, even though they had generally shorter thumbs, demonstrated a generally larger reach envelope that was circular, as opposed to the ellipse young fathers produced (young mothers were somewhere in between). Otten et al suggest this could be due to a more relaxed grip by the adolescent females, which could be due to greater experience with handheld electronics, or that it could also be due to greater joint flexibility among that demographic.

Finally, after reviewing this research and thinking of the fast growing number of thumb controlled handheld devices entering the market, I can't help but wonder what affect they may have on hand/thumb health and effective task performance among frequent users? Time will tell, but the more companies that apply an informed design approach, which by definition must at least include population and task considerations, or be a true human-centered (ergonomics) design, the better it will be for users. 

Bravo to Otten, Karn and Parsons, and their companies, for their exemplary ergonomics approach, and for sharing their data.

Oh, and one last thought: I also enjoyed this research review because I got to use the phrase "Mr. Potato Head" in the otherwise stodgy world of research terminology.

Reference

Edward W. Otten, Keith S. Karn and Kelley S. Parsons, (2013). Defining Thumb Reach Envelopes for Handheld Devices. Human Factors, Published online before print January 23, 2013, doi:10.1177/0018720812470689.

At the time of this writing, this article was avialbe at no charge: http://hfs.sagepub.com/content/early/2013/01/22/0018720812470689.full

This article originally appeared in The Ergonomics Report™ on 2013-01-30.