Do bigger or more monitors increase productivity?

Question

There is a ton of anecdotal evidence that adding more screen estate (more or larger monitors) leads to an increase in productivity, especially for programmers.

For example, where I work, all programmers have either two 20" screens or a single 24" screen. There is no agreement on which is the optimal setup.

The question has been extensively debated by Jeff Atwood and Joel Spolsky:

More usable desktop space reduces the amount of time you spend on window management excise. Instead of incessantly dragging, sizing, minimizing and maximizing windows, you can do actual productive work.

— Jeff Atwood

Debugging GUI code with a single monitor system is painful if not impossible. If you're writing GUI code, two monitors will make things much easier.

— Joel Spolsky

Unfortunately, it seems to me that there is no indication that these effects are real, measurable and properly studied.

• The papers linked by Jeff (and around other blog posts), are all sponsored by monitor making companies (NEC, Apple...)
• The papers disagree on their conclusions: one says that a 30" monitor is better than smaller monitors, the other that there are diminishing returns.
• Even the bloggers do not agree whether increasing the size or the number of monitors is the correct thing to do in order to increase productivity (Jeff thinks that very large monitors are bad, but a three monitor setup works, others say exactly the opposite).

What are some reliable, unbiased studies that describe correctly the relationships between number of monitors, size of monitors and productivity of software developers?
Is there a set up which is proven to work better than the others, or is this an unclear effect that needs more studies?

Answer 1

This is a valid question, and something that I'm partially looking into for my thesis. Based on some initial investigation, Microsoft has done quite a bit of research in that area.

Most of the papers show that increased screen estate and multiple monitor setups can indeed improve productivity by reducing detrimental effects of interruptions. Modality of interruptions seems to be a very important aspect since spatial memory and visual cues are important for reducing primary task resumption times. The paper by Baudisch especially focuses on various ways that multiple monitors can be used, and which ones are most effective for particular use cases.

Here are some paper titles that I have come across so far:

Effects of Visual Separation and Physical Discontinuities when Distributing Information across Multiple Displays (Desney S. Tan1 and Mary Czerwinski) PDF

Partitioning Digital Worlds: Focal and Peripheral Awareness in Multiple Monitor Use (Jonathan Grudin) PDF

Toward Characterizing the Productivity Benefits of Very Large Displays (Mary Czerwinski) PDF

Keeping Things in Context: A Comparative Evaluation of Focus Plus Context Screens, Overviews, and Zooming (Patrick Baudisch) PDF

Display Space Usage and Window Management Operation Comparisons between Single Monitor and Multiple Monitor Users (Dugald Ralph Hutchings) PDF

Using Peripheral Processing and Spatial Memory to Facilitate Task Resumption (Raj M. Ratwani) PDF

Answer 2

There is a study by Microsoft Research comparing a 15" XGA flat panel with a prototype 42" curved screen using three XGA projectors. They state in their conclusion

Users were significantly faster working on the large display. In addition, all but one participant preferred carrying out the tasks on the larger display surface, and user satisfaction measures were significantly better for the larger display.

Another study comparing one 1024x786 (XGA) with two XGA displays found a benefit of dual screens:

We found that multiple monitors helped reduce task time and workload in spite of a learning effect. We also examined how prior multiple monitor usage affected performance and found regular single monitor users benefited from a multiple monitor setting in terms of both task time and workload.

A third study used a 31.5 megapixel display composed of 24 17" displays. They concluded

As the pixel-count is increased (up to 31.5 million pixels) on a high-resolution (96 DPI) display, users will perform faster on some tasks, with more physical navigation and less frustration.

In this study they also make the case that the curvature of the screens is important for such large displays.

Answer 3

Having championed the purchase of five 30" for the equity research officers at Itaú Securities, I have some anecodotal evidence to share, which pretty much invalidates any study regarding 30" monitors.

Here's the deal:

A 30" monitor is pretty large and its distance from the user GREATLY impacts its usability. For a 30" monitor to be usable it needs to be placed on the desktop at about an armslength to be comfortable and productive. For many desks, this distance is an impossibility, especially corner desks where the monitor is catacornered.

Most people when they use an external screen place it just behind the keyboard. This does not work with a 30" monitor, as it is too close and you end up shifting your head up and down and laterally far too much.

This effect is significant enough to invalidate pretty much any study that doesn't take this into account and control it.

At Itaú Securities all five 30" monitors were first used on corner desks with the monitor catacornered. The major factor which determined if it was useable was each analysts body posture when seating at their desk. Those analysts which sat slouched in their chair and leaning back when they were modifying their financial models in Excel, loved the monitor and said that others would be allowed to use it when you pry it from their cold dead fingers. Those with good posture ended up sitting far too close to the screen and since the monitor was catacornered, they were unable to adjust the distance to be comfortable. None realized that monitor distance were the reason for their discomfort. All that sat with good posture abandoned the 30" monitor and return to their crappy 19" monitors. It was only when one of the monitors was relocated to a non-corner desk and placed at the very back of the desk when we realized the impact of monitor distance and comfort.

Anyways, I've never seen this particular issue with larger monitors discussed anywhere with regards so I figured I'd share it here. I'm sure this issue is equally applicable to multi-monitor setups.

Answer 4

As a software developer, I am many times more productive with two monitors than one. I have two 24" monitors at home and a 24" and 19" setup at work (I love my home setup WAY more). The second monitor is where Explorer, SFTP clients, PuTTY, and Excel sit - usually. The majority of my time is spent on the primary monitor with Firefox, Visual Studio or whatever the IDE of the week is, and other tools. I run almost all of my applications full-screen on the primary monitor. Windowed apps run on the secondary monitor.

I also brought in my own leather chair, which is a bit nicer than what the CEO has. It reduces/eliminates my headaches by providing support for my head and therefore also contributes to my productivity. It is also good for the occasional quick nap.

The industry's rule of thumb is to spoil software developers. Give them awesome hardware, toys, freedom to bring in their own office chair, and they will reward you with awesome software, revenue, and loyalty. Bringing in new developers due to turnover and training them is super-expensive. So a few thousand dollars on nice toys and dual-monitors to keep those employees happy is really nice for the bottom line.

I much prefer dual-monitors to a single display no matter how big it is. Two 24" monitors is sometimes too much for me and a lot of it goes wasted in those cases but sometimes I really need to use all available screen space. Two 19" monitors is plenty of real-estate for most people who are doing little more than budgeting, web browsing, e-mail, and IM/Facebook.