Population and Housing Census 2016

The most distinct feature about the 2016 Population and Housing Census was using new technology –tablets using CAPI for data collection. Tonga had been the first country in the Pacific to use tablets for a Population Census.

Statistics Department had planned to take into account the great advantages that associated this methodology. Tablets were able to contribute to cutting down budget cost, and avoid long period taken for data processing phase and not the least enable monitoring of fieldwork simultaneously. Later advantages include institutionalizing across ministries these equipment. There were 600 tablets all together where 100 was donated by UNICEF and 500 purchased by Tonga Government. Out of those 600, 55 Huawei tablets were purchased first and used during the Census Listing and 545 Samsung tablets were later purchased prior fieldwork.

Computer-Assisted Personal Interview (CAPI)

To improve data quality in data collection, The Tonga National Statistics Office (TNSO), in collaboration with the Statistics for Development Division (SDD) of the Pacific Community (SPC), embarked on a new venture to use CAPI technology in their data collection for the 2016 Population and Housing Census.

One the biggest challenge in National Statistics Office is the collection of good quality data – data which is reliable and timely and the TNSO was one of the first countries in the Pacific region to use the technology in their Population Census.

Through the recommendation and technical assistance of SDD, the World Bank Survey Solution application was used for the data collection which involves the use of 600 Android Tablets to be used by Interviewers for the data collections, and Supervisors and Headquarters in their monitoring and evaluation of the data collection.

The Survey Solution software was identified as one of the best tool for data collection due to its capabilities of monitor data collection and data quality. Every completed questionnaire is sent to the respective supervisors to check and either rejects the questionnaire back to the interviewers if there are any errors/inconsistencies or approves the questionnaire if he/she is satisfied with the completed questionnaire. Rejected questionnaires are then verified and corrected by the interviewers with the respective household while they are still in the field, which ensures data quality is collected from the field by the interviewers.

The use of the GPS in the tablets, together with the capturing the photo of each visited house ensures the interviewers are visiting their correct Enumeration Area (EA) as well as proof of visiting the house. With the GPS points, Supervisors and Headquarters staffs were able to plot these GPS points on the map and identify those interviewers who may have enumerated in a wrong EA or had overlapped to other EAs assigned to another interviewer. In summary, improved data quality is attained through correcting survey processes during the field operation.

One of the biggest advantages of using Survey Solutions was the capability to allow for data collection to continue without internet access and this gives more flexibility for interviewers to undertake their work in areas without internet access and only synchronizes the completed questionnaires once they have access to the internet later.

Digicel was the main service provider for the technology which includes tablets for the Pilot, 3G SIM Cards, DATA and other technical assistance in terms of APN (Access Point Network) and daily reports of data usage by the tablet users. There was only one island where the other service provider was used as Digicel coverage was not so efficient.

The Questionnaire was designed using the Survey Solution Questionnaire Designer Tool, which has the facilities to include validations, consistency and logic checks as well as well-designed skip patterns to assist the interviewers in collecting good quality data.

The use of CAPI involves massive planning, preparation, training and testing and this resulted in a number of trainings being undertaken by the staff of the TNSO with SDD in Questionnaire Design and the full process of using Survey Solution Tools.

A listing was undertaken first, which was a good introduction to the technology as it was only for a few questions and the staff, together with about 50 project staff, were able to learn the different processes involved as well as get acquainted to the use of tablets for data collection.

Following the listing, was the Pilot testing and then followed by the Training of Trainers (TOT) of about 100 persons who went through intensive training for two full weeks in the full processes of using tablets. Most of these training sessions were mainly on practical work where they were trained to being an Interviewer, a Supervisor as well as being a Headquarter so they fully understand the different processes and responsibilities involved with these three components of Survey Solutions. After completion of the training, they were then dispatched to their respective areas to train their interviewers before the actual enumeration starts.

FIELDWORK

The hierarchical structure of the tablet system was of clear defined levels and one which ensured that all was cooperating also enabled the SD office to monitor the work of the enumerators and supervisors during the period. Issues that aroused were dealt with according to the hierarchy. The highest level of the hierarchy was the Headquarter which were the officer who did the final approval of the questionnaire form. Second level of the hierarchy is the Supervisors which was the senior officers and teachers who were responsible for the first hand detail checking of the questionnaire forms send to them by the enumerators. The bottom rank was the enumerators who were responsible for enumerating and sending the forms to the supervisors for checking and from there to the headquarters.

The enumerators were divided into groups in according to the location of the block they were to enumerate. Each group consist an average of 6 enumerators who were to report to 1 supervisor. Each supervisor was to report to headquarter and headquarters was responsible for an average of 2 supervisors and respective teams. Every enumerator was assigned 1 census block to enumerate in time using the tablets that was distributed to them during the training period.

The selection of Enumerators and Supervisors was made from primary school teachers. Where possible, the teacher who resided or close to a particular census block was the enumerator of that block. This was an additional asset due to the local knowledge of the teacher about the census block boundaries and its residents.

The enumeration procedures were publicly announced on the first day of the enumeration, the 28th of November, noting that the responsible enumerator should visit all households. This first visit was to update the household listings. The enumerator and the head of the household were to arrange a mutually convenient time for enumeration to take place during the enumeration period, from the 28th of November to the 16th of December 2016. This was to ensure that the enumerator minimized repeated visits to the households and
that the households would be prepared and available to spend enough time to answer the questions. In addition, if a household was not visited, then one of the household members was asked to contact the SD so that it could ensure that the responsible enumerator followed up.

On the other hand, if a household was visited more than once within this period, then, the household should contact the SD for clarification and to prevent over counting by having a household enumerated twice.

A direct toll-free telephone line was established in the head office for teachers to call for clarification on any problem encountered during this process. A CUG (Closed User Group) was set for all tablets to enable the enumerators and supervisors to communicate freely with regards to fieldwork queries.

The census team worked cooperatively to distribute the tablets and other materials such as questionnaire manual, bags, umbrella and t-shirts during the training period for practicing and to use when the enumeration period began. At the end of the enumeration period the supervisors were advised to collect all tablets and accessories of their respective groups and hand in to the SD office to census team who were checking and collecting them.