Question types

Questions in Section 3 —yours— can be of six types. The first three are scales and are the ones that produce an acceptance percentage; the other three are read differently. If you measure eNPS, your questionnaire also gets a Section 4, explained further down.

The six

Why escala_3 uses numbers

The verbal anchors of escala_1_4 are verbatim from the official instrument. For three points no wording is defined, and inventing it would be inventing instrument: the words on a scale assert something about the distances between its values. So the numbers are shown instead.

Which ones produce an acceptance percentage

The three scale types —escala_1_4, escala_3 and si_no— produce an acceptance percentage: what proportion of responses fell on the favourable side. It is the number you will see most often and the only one comparable over time.

Option, ranking and open questions do not have one —there is no order among their answers— so they are read by their distribution or their content.

See Reading a result.

How each is charted

Scales and si_no — A distribution bar with the spread of responses, plus the acceptance percentage. In the historical view, a line of that percentage over time.

eNPS — A three-segment bar per branch (detractors, passives and promoters) with each branch's eNPS beside it. The table also shows the full spread from 0 to 10. In the historical view, a line of the eNPS on a −100 to +100 axis.

Multiple choice — The spread across options. Adds up to 100 %.

Ranking — Counts, never percentages. Its selection is multiple, so counts add up to more than the responses: normalizing to 100 % would draw a distribution that does not exist. That is why ranking stays out of distribution charts and is read in the table.

Open — Not charted. Read.

Section 4: eNPS

eNPS (employee Net Promoter Score) measures how likely your people are to recommend the organization as a place to work. It is not a type you pick when writing a question: you turn it on with the «Medir eNPS» switch at the top of Configuración → Boletas.

Once on, every questionnaire gets a Section 4 with the base question:

¿Qué tan probable es que recomiendes esta organización como lugar para trabajar?

The base question is fixed: it has the same wording for every organization and cannot be edited or deleted, just like Sections 1 and 2. That is what makes your eNPS comparable from one year to the next. Below it you can add follow-up questions of any type, such as «What influenced your score the most?». The eNPS comes only from the base question.

It is answered on a 0 to 10 scale and answers fall into three groups:

The eNPS is the percentage of promoters minus the percentage of detractors. That is why it runs from −100 (all detractors) to +100 (all promoters), and it is not a percentage: it is points. Passives neither add nor subtract, but they do count in the total.

Three things worth knowing:

• It cannot be switched on or off while a study is running. It would change the questionnaire while people are answering it.
• Switching it off deletes nothing. It removes Section 4 from the questionnaires, but earlier results stay; if you switch it back on, the historical series continues.
• Follow-up questions can hang from the base question, such as a «why did you give that score?» shown to everyone. What you cannot do is show them depending on the value chosen.

Like any result, the eNPS is hidden if fewer people answered it than the confidentiality threshold.

Conditional blocks

Any question can have children: questions that only appear depending on how the parent was answered. Most useful with yes/no questions.

Are you satisfied with communication in your area?   → No
   What is the main problem?                         → Lack of information
      What information is missing?                   → [text]

Nesting is not capped at one level: a block can go three or more deep, and every level is captured and published.

Deleting a question takes all its descendants, at any depth. Before you confirm, the product counts them and tells you how many — from the list it is not obvious, because children appear as ordinary questions.

Which type to choose

If you want to follow it over time, use a scale. It is the only thing that produces a number comparable year over year.

If you want to understand a why, use a conditional block. A yes/no that opens a reason question gives you the metric and the explanation, and only shows up for people who have something to say.

Use abierta sparingly. It is not charted and has to be read one by one. As the closing question of a conditional block it works very well; as a standalone question to the whole company it produces a thousand paragraphs nobody will process.

Before adding questions

Remember that Section 3 questions are yours but not free: they lengthen the questionnaire, and a long questionnaire lowers participation — which is what decides how many results you get to publish. See Announcing the survey.