10 steps of quality control
Providing premium online methodology

10 steps of quality control

Premium online methodology
KNOTs Research has been working since 1996 to develop and manage active access to households worldwide. Building on our technical knowledge pioneered in the Japanese market, we have been active in Europe since 2001. All our work is carried out in accordance with the strictest market research professional standards codes and guidelines and we only conduct surveys for the purpose of marketing research and intelligence. We maintain and update full profile information for all our members on a regular basis and also actively recruit new members to join. Keeping the membership database fresh and engaged has resulted in high response rates for every survey – typically, depending on the complexity or scarcity of the sample, or the sensitivity of the subject, this will be between 12% and 35%. We have defined and implemented a clear set of rules to ensure that quality is our highest priority and support this through regular contact with our members. Currently, we have access to more than 40 million households in over 70 countries – and manage the relationship with global consumers on a daily basis.

10 steps of quality control
1. Maintain database – reliable and up to date 2. Confirm respondent’s identity 3. Screen with strict procedure 4. Select accurately the right sample 5. Inspect and test questionnaire thoroughly 6. Use self-completion method 7. Create user friendly questionnaire 8. Monitor feedback extensively 9. Check questionnaire responses rigorously 10. Offer credible reward system

1

10 steps of quality control

Database
A large household access requires a wide range of sources for recruitment mainly to ensure that KNOTs Research avoids systematic biases inherent to certain sources. KNOTs Research recruitment is based on large portals such as Yahoo, T-Mobile, Google, in order to reach as varied access as possible. In addition, we target certain demographic segments that may be difficult to reach using single sources (e.g. portals for mothers with small children, people above 60 years old, people with unique occupations or hobbies). Every day the KNOTs Research database is refreshed with new people. Data relating to age is automatically updated and we regularly encourage and motivate people to update their own personal information through easy to use online facilities. Anyone who has not updated his or her personal information after 24 months is automatically removed from the database and the system meticulously checks and cross checks data to identify and eliminate ‘double identities’. To make sure that our consumer access accurately reflects the make up of the national population, demographic representation is checked and updated every quarter.

2

Respondent’s identity
KNOTs Research holds key information about the people who we invite to take part in research surveys. For example, we know their age and gender, where they live and what kind of products they own. However, we always check and revalidate critical data in the ad-hoc questionnaire. So, if for example we invite someone from the database who is male, 37 years old and the owner of a specific brand and model of camera to be part of a survey, then we re-confirm these critical pieces of information through the ad-hoc questionnaire and screen out any potential respondent where there’s a discrepancy in the information provided. This double confirmation method helps us to remove unsuitable respondents and have maximum confidence in the validity of the results we get back from research.

3

10 steps of quality control

Screening
Respondents have no idea which target group we are looking for and which brand or other variable are needed to complete the survey, so the screening process is highly reliable. If a client asks to invite people who own or intend to buy a certain product or brand, then we insert the required products or brands among a long list of other options. Only those specifically choosing the required product or brand will qualify to continue to the main questionnaire. And once a respondent has been screened out (i.e. disqualified), there is no way back. People can’t go back and change answers and this helps us to ensure high response quality.

4

Sample
Anyone invited to participate in a study is provided with a new unique ID, which is valid only for the specific study in question. We have ID checks to monitor the respondent’s entrance the survey – the ID checks also ensure that the person invited enters the questionnaire only once. Clients can decide whether respondents from previous surveys should be invited for a second time or not. We also have an automatic rotation feature which enables the KNOTs team to recruit the sample randomly, so that no ‘fixed’ sample is repeated for surveys in similar industries or markets. All potential respondents start the questionnaire with screening questions, based on the client’s specifications (age, gender, brand and product ownership, etc) and only those who qualify based on these requirements are allowed to continue to the main study.

5

10 steps of quality control

Questionnaire
Each questionnaire is tested thoroughly and undergoes a careful proofreading procedure to make sure that ‘respondent friendly’ language is used. This avoids confusion or possible misunderstanding. Questions are asked objectively to prevent respondents feeling pushed towards certain answers. Without this attention to the tonality and phrasing of questions, the validity of answers could be significantly decreased. Questionnaires are translated by certified professionals. Each translator is required to live in the country where the study will be conducted to ensure that the translation can be fine-tuned within the relevant cultural context.

6

Self completion
All questionnaires are programmed online by a dedicated KNOTs Research team who apply a ‘logic flow’ test and so ensure that respondents can independently complete the questionnaires easily and smoothly. The order of the answer options can be shown randomly to respondents to avoid any ‘list order effect’ and response bias. For example, a list of different products, brands, concepts, ads or TV commercials will be presented to each respondent in a different and random order. Our programming team always makes sure that any ‘error messages’ carry simple and clear instructions on how to complete the questionnaire and navigate the interface.

7

10 steps of quality control

Respondent friendly
Each question appears on one single screen with no or little scrolling, so the respondent can focus on the question in hand. If there’s any clarification needed about any of the questions, the respondent can click on the ‘help’ button that appears on each screen and one of our team will get back to the respondent within a few hours. For any questionnaires lasting longer that 12 minutes, there’s an option to take a break and continue with the questionnaire at a later time. All the respondent needs to do is click ‘I want to interrupt this survey’ and then rejoin by clicking back on the link in their invitation email. He or she then rejoins the survey from the point where they took the break and all the input up to that point has been saved.

8

Fieldwork monitoring
The automatic closing feature calculates the feedback quantity per member or attribute and ‘closes’ the survey once the ‘attribute quota’ has been reached. Any respondent completing the questionnaire in less than half the average time required is automatically taken off the ‘completed respondent list’. This person is then also deleted from the KNOTs Research access list. The top liner feature allows us to view respondents’ answers in real time. Consequently, we can determine together with the client at any moment whether to extend or shorten the fieldwork timing or to continue to invite extra members to participate.

9

10 steps of quality control

Response check
For the closed questions, once fieldwork has started, we run a check on all completed interviews to double check validity of the questionnaire structure and flow. It’s only when we’re absolutely sure that everything is working well that we invite a bigger sample to complete the survey. For the open questions, we go through the results and eliminate the entire interviews of any respondent who has not provided ‘valid’ answers. These respondents are then no longer invited to participate in any future study.

10

10 steps of quality control

Rewards
KNOTs Research has designed a unique rewards system based on the accumulation of points that can be redeemed against gift vouchers for leading department stores around the world. We let our members use their rewards to buy or put towards the things they really want. Although the rewards could be sent by email, we employ a strict policy of sending rewards to the real postal addresses of our members. This helps us to perform another check on members’ details and reconfirm identity and some of the personal data. We have found that this system ensure much higher levels of commitment from our members, resulting in higher than average response rates and much quicker response times.

Next Steps
Our quality control approach delivers a constantly growing level of access to households worldwide. Our methods of recruitment and ability to target specific segments and hard to reach profiles are constantly under review and development, keeping pace with the reality of how people interact with technology and communication. Through the processes which we have embedded in our operations and the level of direct control we exercise, we are well placed to check quality at every step of the way. We see that our members take time to complete the survey questionnaires and remain committed to providing thoughtful, complete and valuable insight into the surveys we conduct for our clients.

KNOTs Research Neptunusstraat 31 2132 JA Hoofddorp The Netherlands t: +31 23 554 3 554 f: +31 23 554 3 553 e: info@knotsresearch.com www.knotsresearch.com