September 28, 2021
In human factors, we often talk about the importance of completing a task correctly—without use errors. However, an element of completing a task that is seldom talked about, yet quite important, is the time it takes to complete a task. If a user can administer medication using an injection device without use errors, but can only do so if they spend 30 minutes to an hour on the activity, that isn’t considered a success. Not only should users be able to complete tasks correctly, but they should be able to complete tasks efficiently.
Identifying each use step within a process
In this blog, we consider a task to be one use step within a process. For example, removing the cap on an autoinjector is one task within the activity of administering medication.
What is the appropriate time for a user to complete a single task? Based on working memory, it should take about 10 to 15 seconds to complete a single task within a process.
This might be easy for users who are familiar with a product and its workflow (e.g., having injection experience) because they could focus on completing the task and likely would not need to read instructions. However, for users who are new to a product (e.g., injection-naïve), they not only need to complete a task, but they need to first read instructions and understand what is being asked of them. This is where things get complicated, because we must consider three elements of completing a task for new users: reading instructions, understanding the task (based on the instructions and device’s design), and performing the task.
Clear and timely instructions
According to AAMI HE75, ”Each step should be written so that the user can read and complete it in about 15 seconds,” which aligns with the expectation that completing a single task should take 10 to 15 seconds in order to better align with average working memory. Operating under this assumption, this means that users have 15 seconds to read an instruction, understand the task, and perform the task. Or, to represent it as an equation:
Treading + Tunderstanding+ Tperforming = 15 seconds
Based on this equation, we have a budget of 15 seconds per task during which a user must read, understand, and perform the task.
In many cases, understanding a task happens in tandem with reading the instructions. So let’s assume that Tunderstanding = 2 seconds to account for any quick processing after reading the instruction. That means we are left with:
Treading + Tperforming = 13 seconds
Let’s also assume that physically performing the task takes about five seconds, on average for a simple interaction. We are now left with:
Treading = 8 seconds
This means that, based on our assumptions, users have eight seconds to read an instruction before understanding it and performing the associated task. How much can we expect someone to read in eight seconds?
Here are a few key factors related to readability to help us determine this:
- A single step’s instruction should be one sentence, and a maximum of 25 words in a single sentence.
- According to the FDA’s guidance on patient labeling, instructions should be written for no higher than an 8th grade reading level—a 6th to 7th grade reading level is preferable.
- A 6th grade reading speed is 185 words per minute, or ~3 words per second.
Therefore, if an instruction contains no more than 25 words, it will take a little over eight seconds to read, which is the hypothetical amount of reading time we had available for our task:
8 (Treading ) + 2 (Tunderstanding ) + 5 (Tperforming ) = 15 seconds
From this exercise, we can deduce that a single-sentence instruction that is written for a 6th grade reading level and contains 25 words or fewer will enable a user to read, understand, and perform the associated task within the 15-second timeframe recommended by AAMI HE75.
Based on our formula, the three variables are inversely proportional to one another: as one increases, the others must decrease to stay within the 15-second timeframe suggested in HE75. As such, how you distribute your 15 seconds across the three parameters per task can vary based on your needs while developing the instructions. Considering working memory, time to complete tasks, and reading speed enables you to think about how the tasks and associated instructional steps can vary throughout an entire user interface (e.g., some will take more time to complete, or are more complex and will require more detailed instructions) and, relatedly, how these 15 seconds will be distributed across a given task.
Readability and effective patient labeling
Although this equation is a useful means of objectively assessing the parameters of an instruction, the most essential test of readability and successful instructions is not to simply design by formula. Rather, it is to evaluate the instructions with real users throughout the design process and, ultimately, validate the instructions with real users when the instructions become production-equivalent. To cite the FDA’s guidance on patient labeling, “the reader must actually read the text to determine if it is readable.” Similar to readability formulas (e.g., Flesch-Kincaid) leveraging a formula like this can provide a good starting point and help structure the design approach with useful guideposts. However, it is ultimately the observation and validation of the users’ ability to read and understand the instructions that is the most important measure of the instructions’ quality.
Alexandra Benbadis is Senior Human Factors Specialist at Emergo by UL’s Human Factors Research & Design unit.