Panels Online

FAQ and Myths About Online Research

We thought about the doubts and barriers that exist with regards to doing online research and online data collection. Here are our answers to some of the most important frequent questions ...


1 - Which population are we talking about and is online data collection representative?

We are talking about the online population of Brazil, which consists of about 50% of the total population (April 2014), a number that keeps increasing fast. Our aim for CONECTAí is to be representative for this internet population. We control the profile of the panel continuously. Hence it is possible to use stratified sample quotas and apply methods such as weighing and ponderating data if necessary.

If you need your research data to be representative for the whole of the brazilian population, other data collection methodologies should be used instead of or together with the online methodology

2 - How can I know if the obtained sample is representative for my target aucience?

Yes we can do research with the general public (general internet population) or with specific target audiences, with sub-samples based on a very wide variety of variables: demographic, financial, consumption-related, behavioral, educational, atitudinal and other variables.

A maioria dos perfis de consumidores está presente na internet.

For example: specific profiles such as doctors, executives, supermarket frequenters, credit card users, smarft phone users, people with post graduation degree who thave travelled during the last 6 months... and much more.

How do we know these profiles? Because we continuously collect profile information on all panel members via special questionnairs and through normal surveys. We also control the sample during fieldwork.

And what if we don't have enough panel members in the CONECTA panel for a specific request? In this case we work together with partners to complete your sample as requested or recruit the specific profile you need for your project into our panel so we can increase the number of completes.

3 - How do I know that the panel members are who they claim to be? How to know if the respondent is really the person I want to get opinions and ideas from?

This is one of the many big advantages of working with an online research panel instead of using other methodologies;

  • Even though participation is entirely voluntary, there are various checkpoints at the moment of registration of a new panel members and we collect profile information on the person from the start on (name, CEP, CPF, age, ...). Further more consistency checks on profile and survey level are carried out regularly:
    • Consistency between propfile and answers to checkpoint questions in surveys, consistency between answers within survey, consistency of answers between surveys.
  • It is crucial that surveys are well elaborated to allow these checkpoints, including blind filters and confirmation questions.

4 - Why do people participate in a research panel?

  • To share their opinions and express what they think
  • To know what others think about certain topics
  • To be part of na engaged community of people
  • To winn prizes
  • To make a diference and encourage change and improvement
It’s important to create a feeling of importance for the panel member, for him / her to feel part of a valuable comunity and for example even be able to invite friends to take part as well...

5 - How do panel members get remunerated for their participation in CONECTA’s surveys?

  • Our panel members accumulate points which they can afterwards exchange for a very wide variety of prizes or that they can donate to charity.
  • On top of that for each survey answered the panel member participates in a lottery of a pool of prizes.
  • The items of the lotteries and number of points depend on the topic, complexity, duration and other characteristics of the survey.
How to avoid ‘survey professionals’?
  • Limiting the number of survey invitations and participations per month
  • Limiting the number of surveys concerning a topic within the same area
  • Controls to avoid duplicate accounts (CPF, profile checks, ...)

6 - How to know the participation rate in my surveys?

Surveys realized in online access panels, like CONECTAí have a response rate that varies between 20% and 50%, depending on a number of parameters such as; theme, duration and format of the survey and the incentives given.

We also know that on a yearly basis about 15% of the panel members drop out due to a variety of reasons, which requires us to replace a part of the members continuously.

  • Our internal ‘teaser surveys’ are a good example. For example the CONECTA study about Carnival 2013 was responded by 1.374 panel members out of a total of 3.000 people invited. Fieldwork was wrapped up in just 3 days; our survey concerning the political debate of Cracolandia was sent to 2.600 panel members of which 1.233 participated in only 3 days time as well. These surveys are important for our panel members to express their opinions on social interest topics, get feedback in the form of the results of the survey and even see their opinions have repercussion in the media and inspire change in society

7 - Can I transform an offline research project into na online project?

Yes you can, however some aspects need to be taken into account and diferences are inevitable.

  • The questionnaire needs to be adapted for the online environment (tone of voice, format, duration). This is a good opportunity to improve the questionnairs in fact!
  • The sample will probably be different.

If the original survey was conducted using lists of telephone numbers, the ideal approach is to migrate to conducting the survey to online using the same lists, just adding the emails of the contact persons.

In case a comparison of the results between offline and online is required, the ideal is to run both studies paralel for a little while. The investment will be recuperated later in time when the transition to online is more complete.

8 - Which parameters determine the price of an online data collection project?

  • Survey length
  • Incidence rate of the product category
  • Demographic incidence rate (of the target audience)
  • Complexity or sensitivity of the topic
  • Field work duration

Be aware: wrong briefings can results in wrong price quotations...

9 - When to do or not to do a research project online?

Most of the studies any organization wants to do can already be done online.

A internet está crescendo de forma acelerada e a maioria dos perfis de consumidores está presente na web.

Some exceptions are:

  • Studies that require pure a probabilitsic sample representative for the whole brazialian population, including rural zones and the lower social classes (DE)
  • Studies with very long questionnairs.
  • Studies that require observation of the respondent during the taking of the survey or that require live interaction with the respondent.

10 - What are the advantages of working with an online research panel?

  • Modern language
  • The computer / internet is the current comunication channel of choice
  • The respondent can participate when and where he prefers
  • Shorter surveys because we already know the panel member (no need for repeat information)
  • Higher response rates -> faster field work
  • Geographic reach
  • Speed
  • Cost
  • People are less inclined to open the door or talk to strangers
  • Refusal rate to participate in telephone surveys increasing
  • More and more people no longer have fixed phone (more difficult to reach and higher costs) + people don’t answer cell phone when number is unknown
  • Easier to present stimulus / audiovisual materials
  • No interviewer bias and more question types and options available
  • Different ways of asking questions
  • More access to difficult profiles with lower incidence rate

Your turn to ask


Internet users have a different consumption profile from who does not use the internet?

  • There are many offline and online factors influencing and determining a person’s behaviour. The access to the information available on the internet might infleunce people’s decisions, but that is part of reality. Factors such as social class and region for example have na influence a lot bigger than the fact if a person has internet access or not.

An internet user participating in online surveys has a different profile than those who don’t participate?

  • The same can be said for people that participate in online telephone or face to face surveys ... If the bias exists it exists for all methodologies, offline as well as online.

people lie on the internet?

  • Maybe the fact of being ‘anonimous’ (internet) induces a person to lie more, however at the same time it is more likely that the person will answer more truthfully to questions as no ‘social pleasing’ takes place.
  • Studies show that for sensitive subjects the online methodology is more effective. At the same time purchasing intentions declared in online surveys seem to be more realistic!

Only young people, men and high social classes are represented on the internet?

  • Currently, in Brazil, half of the internet users are women, over 22% are 50 years of age or older and half the internet users can be categorized as social class C.

People don’t participate in online surveys that last over 5 minutes?

  • It is a fact that short surveys are more attractive and easy to respond, however depending on the communication and relationship with the panel member, the topic of the study and the incentives given, people tend to participate to studies that need more time to complete.
  • Best practices indicate that 10-15 minutes is a good survey length, whereas 20-25 minutes should be considered as the maximum length for online surveys. Longer studies are possible but not recomended.

In na online access panel it are always the same people that are responding to all surveys?

  • The number of surveys each panel member can participate in, as well as the themes of the studies, are controlled and limited per month in order to avoid this.
  • On the other hand: it is interesting to track a person’s behaviour over time (trackers, migrations, ...)

Newsletter Subscription

Go to top