Say you’re filling in a form. One answer is already selected, and if you accept it you can carry on straight away. Another answer is there too, but you need to notice it and decide to change the selection.
You still have a choice. Yet saying the form simply offers two answers leaves out how it shows them. The preset matters to understanding the choice, even before you know whether it changes anyone’s answer.
You can look at that design without assuming it controls you. The useful question is how its paths fit the task you want to get done.
This module is for working adults judging how a tool is likely to affect a particular task in a particular setting.
What you’ll get from this module
In this module
By the end of this module, you’ll have a clearer way to judge what a tool makes easy, hard or automatic. Design theory, a research review and field studies show why these differences deserve attention, and why results depend on the setting. You’ll link those findings to everyday choices, pick one situation and finish with a written judgment or a clear open question.
Ask Jing
Become a Jingius.
How design shapes the paths available to you
Design theory: an available action and a visible cue are different.
Donald A. Norman’s 2013 book explains an affordance as a relationship between an object and what its user is able to do. A chair that one person can lift may be too heavy for another. Whether it can be lifted depends on both.
Norman separates those possibilities from signifiers: cues that show where an action should happen. A visible circle on a screen can show where to touch. The cue and the ability to do the action are different parts of the design.
This difference helps separate three questions that are easy to mix up. Does an action exist for this user? Can the user see it? Does the setup actually block it?
Why it matters: A hard alternative is different from an impossible one. A prominent cue is different again: it points toward an action without showing how strongly it changes behavior. Norman gives a way to describe these differences before trying to explain their effects.
Read the design chapter on affordances and signifiers.
Research review: defaults influence choices, with substantial variation.
A default is the option already selected when someone reaches a choice. Jon M. Jachimowicz and colleagues’ 2019 review combined 58 sets of results from 55 studies. This kind of review, called a meta-analysis, looks at results together to see the overall finding and how it varies.
Overall, presets had a large influence on choices. Most results showed choices shifting toward the preset. Several found no statistically significant effect, and two found a shift the other way. So the overall result supports taking presets seriously, while expecting their influence to vary.
The review also looked at possible reasons. Researchers rated whether each study’s default seemed to suggest a recommendation, called endorsement, or to stand for the way things already were, called endowment. Effects were larger in studies rated as having those features. Comparing studies this way doesn’t show why every preset works.
Why it matters: The important difference is between spotting a default and predicting its effect. Seeing which answer is selected tells you what happens if nobody changes it. It doesn’t give a reliable estimate of how many people will accept it in a new setting.
Read the review of when and why defaults influence decisions.
Randomized field experiment: practical help changed an available path.
A choice can also depend on the work it takes to carry it through. Eric P. Bettinger and colleagues’ 2012 study looked at help with the Free Application for Federal Student Aid, or FAFSA, in the United States.
The experiment took place in 2008 among eligible lower-income families using H&R Block tax offices in Ohio and the Charlotte area. Families were assigned at random to one of three groups. One got help filling in the application plus information about aid. One got the information alone. One got a brochure, as a comparison group.
The analysis included 868 children in grade twelve who relied on their parents financially. In this group, the combined offer raised first-year college attendance by 8.1 percentage points, from 34.2% to 42.3%. Information alone showed no benefit for attendance. But that estimate was too uncertain to treat the result as proof of no benefit.
The help combined making things simpler, hands-on help and information. Its result belongs to that combined offer. The experiment doesn’t single out a number of steps, or one kind of help, as the cause.
The useful finding is that having an application available and having support to finish it can lead to different results. Knowing about an option still leaves a person with work to do. The help changed the practical route through that work.
Why it matters: For judging a tool, this gives effort a wider meaning than the number of clicks. The knowledge needed and the help available also belong in the description of a path.
Read the field experiment on financial-aid application assistance.
Randomized field experiment: a default effect wasn’t detected in one specialized setting.
The positive findings don’t mean every preset produces a shift you can detect. Åsa Löfgren and colleagues studied environmental economists signing up for a 2008 conference in Gothenburg, Sweden. They could make a real payment to offset their flight emissions.
The October 2009 working paper reports that people were put into three groups at random. One started with not paying selected, an opt-in setup. Another started with paying selected, an opt-out setup. The third had no preset. The analysis included 240 complete, valid cases after some were left out. Changing a preset took one mouse click.
Offsets were bought in 39.33% of opt-in cases, 43.24% of opt-out cases and 46.75% of cases with no preset. The overall difference was not statistically significant.
This experiment didn’t show a default effect among these people. It also didn’t show that the three setups had the same effect. The researchers changed the presets at random, but didn’t assign people to be experienced or not. So the result doesn’t show that experience caused the missing effect.
Why it matters: The lesson is to leave room for an influence that is weak or not picked up. You need evidence about the setting in question to judge its size. Confidence in your own experience can’t give you that evidence.
Read the 2009 working paper on paid flight-offset choices.
Research using employee records: an initial choice and a lasting outcome can differ.
Even when a design changes a first choice, later events can change what it is worth. James J. Choi and colleagues’ December 2024 working paper looked at automatic retirement-saving policies at nine United States firms.
The researchers compared employees hired before and after the policies began. They followed how much people paid in, and took account of job changes and withdrawals. They also counted employer payments that employees lost because they left before fully earning them. To estimate longer-term results, they assumed people would behave the same way in later jobs with similar savings plans.
Their model estimated a lasting rise in savings of 0.6 percentage points of income for automatic enrollment alone. Adding automatic increases in payments to plans that already enrolled people automatically gave an estimated rise of 0.2 percentage points of income. These estimates mix observed behavior with assumptions about the future. They aren’t lifetime results measured for everyone.
The analysis shows why signing up and lasting savings need separate attention. People can later change what they pay in, leave an employer or take money out. So an early shift doesn’t tell the whole story of what is left later.
Why it matters: This takes the question beyond the easy choice itself. A tool may help with one step toward a goal, while the final result still depends on other actions and arrangements.
Read the December 2024 paper on automatic savings policies.
What this means for you
Judge the design against a task you actually care about. A design has a “tilt” when it favors one path over another. Describe that difference, then work out whether it matters for your goal.
Separate what is possible from what is easy to notice
Start with what you want to get done and who needs to use the tool. Describe the other options open to that person, including any skills or help they would need.
Then separate an action that isn’t there from one that is just harder to find or finish. If the other option exists but needs extra know-how, write down that burden. If you don’t know whether the other option exists, write that down as an open question.
Say a made-up sharing tool prominently offers to send a document to the whole team. You want to send it to one colleague. Finding a way to send it to fewer people answers a different question from finding out whether the tool allows that at all.
A hidden route may still fit your goal once you know where it is and what it takes. A route that isn’t there leaves that goal unmet with this tool. Keep those two judgments apart when you describe the obstacle.
Compare the practical work on both paths
Ask what the tool makes effortless and what it makes effortful. Include time, steps, know-how, help and costs. Compare the actual options, rather than calling the whole tool easy or hard.
The application-help study shows why this needs more than a count of clicks. A visible option can still need information or help the user doesn’t have. Write down what the person would need to finish each route.
In the made-up sharing example, the whole-team option is right there. Sending to one colleague might mean finding another setting. If you haven’t found it, write down that uncertainty, rather than assuming the extra work is trivial.
Describe the default before interpreting it
Write down what happens if the user changes nothing. Then ask what kind of user or task that preset seems to assume. Keep what the preset does apart from your guess about the task it was designed for.
A preset may look like a recommendation, but leaving it as it is doesn’t show what a person prefers. You can compare it with your goal without assuming everyone who accepts it wanted the same thing.
For the made-up sharing tool, what you see is that the ready-selected route sends to the whole team. Your reading is that this route suits a task for the whole team. Your task, for one colleague, may need another route.
Ask who benefits, and what happens afterward
Ask who profits from each easy path. Here, profit means any benefit, including benefits other than money. You may benefit too. Write down the benefits and costs you have evidence for, and leave the rest unknown.
An easy path alone doesn’t show what a provider intends. A benefit can be shared. If a conclusion about intent would change your judgment, you need evidence about that case.
Also name the outcome and the length of time you care about. The savings research shows why finishing an early step isn’t the same as getting the later result. Describe what still needs to happen after the easy choice.
In the sharing example, sending a document is the early step. Getting it to the right colleague is your goal. A whole-team default may make sending easy, while you still have to judge whether its choice of people fits the task.
Decide how much the difference matters, then stop
Give the design’s tilt less weight only when three conditions hold together. The other options are truly open to this user. The user understands the options and knows what they want. Evidence about the situation in question shows differences too small to matter for the result.
Each condition answers a different question. Can the person actually use another route? Do they know which route fits their goal? Is the difference in results small enough to set aside? Having a choice doesn’t answer all three.
If evidence is missing, leave the size of the influence unknown. Knowing a tool well doesn’t settle this. Nor does one study that fails to find an effect. A clear description of the design can still be useful while its effect on behavior stays uncertain.
Aim this exercise at a design difference that could matter to your task. Once you can explain why it matters to your goal, or name an important gap in what you know, stop. A favored path isn’t, on its own, a reason to give up a tool you need.
How about your daily life?
Where would understanding a tool’s paths help you make a judgment? You might think about an application form, a sharing choice or a setting you usually leave as it is.
What result are you trying to get? Which part feels effortless, and which other option would you need to understand better?
Write down a situation you would like to look at. One useful idea is enough.
Situation: ____________________
What I want to understand: ____________________
Choose a situation with a particular tool, task and user. Carry it into the exercise below. Choosing it doesn’t mean anything needs to change.
My chosen situation: ____________________
Read the tilt in one situation
Start with your chosen situation. If you came straight here, choose a tool you use for a particular task, and a design difference that could matter to that task. This is a written judgment.
- Name the goal and context. Write who is using the tool, what they want to get done and the length of time that matters.
User, task, outcome and time: ____________________ - Answer the four questions. Describe what you see, and mark missing information as unknown.
What does this tool make effortless? ____________________
What does it make effortful? Include steps, know-how, time, help, costs and anything actually blocked. ____________________
Who profits from each easy path? Include benefits, in money or otherwise, that you have evidence for, including yours. Keep possible motives apart from benefits you have seen. ____________________
What are the defaults, and what do they assume? Say first what happens if nothing is changed, then label your reading of the user or task they assume. ____________________ - Check alternatives and follow-through. What can this user actually change, in these conditions? What still needs to happen after the first choice to reach the result you named?
Available alternatives, needed support and later steps: ____________________ - Make a judgment about this situation. Explain which design difference matters to the task and why, or write down what you still need to know.
Give the tilt less weight only when all three conditions hold. The other options are open to this user. The user understands the options and knows what they want. Evidence about this situation shows differences too small to matter for the result.
If evidence is missing, leave the influence unknown. Knowing the tool well, or finding one study that found no effect, isn’t enough.
Assessment or unanswered question: ____________________
Ask Jing
Become a Jingius.
Your next step
You have finished when you can describe the paths, relate them to your goal and name any uncertainty that could change your judgment. Stop there, rather than making up an answer. Don’t give up a tool you need just because its design favors a path.
A note on the evidence
The evidence combines design theory, results across studies and experiments in particular settings. An average effect across studies can sit alongside no effect found in one setting. Apply each finding to the result it measures.
The application study couldn’t include everyone who first agreed to take part, because some records it needed never arrived. The information-only group was also smaller, which made its attendance estimate less precise.
The savings paper uses a model to estimate what happens beyond the jobs it looked at. It doesn’t measure the full change in a household’s savings and debts.
How we prepared this module
This module brings together design theory, a review of default effects, two field experiments and research on savings policies. A made-up form and sharing situation show how to compare paths. The exercise turns those ideas into questions about one tool and task.
Sources
- Donald A. Norman. (2013). The Design of Everyday Things. New York: Basic Books. Revised and expanded edition; chapter extract, “Affordances” and “Signifiers,” pp. 11–14.
- Jon M. Jachimowicz, Shannon Duncan, Elke U. Weber and Eric J. Johnson. (2019). When and why defaults influence decisions: a meta-analysis of default effects. Behavioural Public Policy, 3(2), 159–186. Published article; data collection, p. 163; results, pp. 164–166 and 174–175; “Limitations and future directions,” pp. 177–179.
- Eric P. Bettinger, Bridget Terry Long, Philip Oreopoulos and Lisa Sanbonmatsu. (2012). The Role of Application Assistance and Information in College Decisions: Results from the H&R Block FAFSA Experiment. Quarterly Journal of Economics, 127(3), 1205–1242. Published article; “The FAFSA Experiment,” pp. 1211–1217; Table III, p. 1226; results, pp. 1227–1229.
- Åsa Löfgren, Peter Martinsson, Magnus Hennlock and Thomas Sterner. (October 2009). Does experience eliminate the effect of a default option? – A field experiment on CO2-offsetting for air transport. University of Gothenburg, Working Papers in Economics 391. Author-uploaded working text; “Experimental design” and “Analysis,” pp. 4–8; Table 1, p. 12.
- James J. Choi, David Laibson, Jordan Cammarota, Richard Lombardo and John Beshears. (19 December 2024). Smaller than We Thought? The Effect of Automatic Savings Policies. Working paper hosted by the American Economic Association for its 2026 meeting. December version; methods and findings, pp. 2–6 and 9–18; Table 1, p. 38.
