What actually makes a good survey? (According to 216 of them, anyway)

For 15+ years, our platform has adapted to every imaginable survey need. In addition to all possible kinds of spatial planning projects, it's hosted surveys on topics such as wildlife encounters, kids' preferred outdoor play places, and the effects of perceived darkness on spatial quality.

Clearly, a core strength of our platform is its adaptability to different contexts.

The secret lies in flexibility: Maptionnaire doesn't chain you to a pre-defined structure. We know, however, that total creative freedom can be intimidating. Which of the 30+ elements work best? How long is 'too long' for a survey?

To answer this, we sponsored a geoinformatics study to analyze how survey design impacts completion rates and GIS data quality. By auditing 216 Maptionnaire surveys, the research identified specific indicators of high-quality data. For instance, respondents who zoomed in beyond the default view demonstrated greater attention to detail, resulting in significantly more accurate spatial data.

How long should a survey be?

Unsurprisingly, the researcher found that the longer the survey, the poorer the completion rate. Surveys that took more than 15 minutes had significantly lower response rates. 

However, it's also not a good idea to cram your questions into a couple of pages to create an illusion of brevity ("It's just a 3-page survey!"). People really do not like to scroll. Surveys with long pages lost the respondent's attention a lot faster. 

So what should you do then? Compile your potential questions, then refine them to focus only on the essentials. Organize these questions across different pages, only a couple per page. 

Visual and branding choices did not appear to have a significant effect on completion rates. However, simplicity in design was found to be somewhat influential in predicting data quality.

Does it matter which maps I choose?

Every Maptionnaire subscription comes with basemap library, and it's also possible to use your own (e.g. hosted on a city server). Clients often prefer to feature a selection of maps in their survey, thinking that respondents will find it easier to orient if they can switch between them. The research does not support this – it found that using multiple basemaps with different visual styles affected the quality of data negatively.

You can also upload map overlays (like shapefiles, GeoJSONs, or GeoTIFFs) to provide respondents with richer spatial context. For example, a map overlay in a mobility-related survey can highlight existing transport networks. The research supports the smart use of map overlays when they are directly connected to the question. It can help respondents give more accurate answers.

However, remember that you can also go too far with detailed overlays. You don't want the respondent to feel that the survey takes a lot of concentration or find out that their cell phone is not be able to load the maps smoothly.

What makes a mapping question easy for respondents?

The Map point is by far the most commonly used mapping element in Maptionnaire surveys. It also seems to result in better data quality than requiring respondents to draw lines or polygons. This might be because drawing areas and routes requires more steps than placing a pin.

Again, restrain should also be applied to the number of mapping questions on one page (even if they are all Map points). 

Another interesting finding was that asking follow-up questions in a pop-up seems to have a positive effect on the accuracy of the pins placed. Surveys that used more pop-up windows made respondents zoom in more often. The hypothesis is that this is because pop-up questions give respondents more time to focus on the location.

The subject matters 

As a final note – not all topics engage equally. Mobility and transportation surveys see the highest completion rates because they are universally relatable and trigger strong emotions. Conversely, while abstract topics like 'cohesion' attract fewer participants, those who do engage provide significantly more detailed and high-quality insights.

Do's and Don't's 

1. Keep it short ✔  

   Aim for under 10-15 minutes. Write down all that you want to know/ask and then distill them into the essential questions. Merge questions and ruthlessly delete ones that are not must-haves.

2. Split questions between pages ✔ 

   Limit the number of questions and mapping tasks per page. Keep layouts clean and simple. It's better to separate questions from busy pages to separate pages.

3. Don’t use multiple different basemaps in the same questionnaire ✔

   Using a different background map on every page might be visually overwhelming and can distract the attention of the respondent.

4. Smart map overlays can have a positive effect on data quality ✔

   Adding your own map overlays to the questionnaire can help respondents to better understand the spatial context of the questions. 

5. Ask respondents to pin instead of drawing lines (whenever possible) ✔

   Even if you want them to think in terms of a larger area, a pin is often sufficient for delivering the same information as area drawing.

6. Don't forget the pop-up(s)! ✔

   Adding a couple of follow-up questions in a pop-up makes the responses more accurate. Remember to not overdo it though!