Pacific HIES Toolkit

HIES Timeline

From planning to dissemination : key stages during the 3 year timeline.

(click on each section in the timeline to expand)

12 Months – Year 1 to Year 2

The Pacific HIES Toolkit - Planning of HIES

This stage covers initial work planning and design of the survey instruments, survey governance, resource and budget allocation and survey governance.

• Planning survey activities (Month 1, Year 1)

  International recommendations advise to conduct HIES every 5 years; due to funding and availability of resources issues, however, this periodicity might not always be possible. Most of the countries and territories in the Pacific region conduct HIES every 5 to 10 years; it is recommended not to exceed 10 years as the lack of updated data will become problematic after such a long-time span.

• Design and Build (Month 2, Year 1)

  During the survey planning phase, it is recommended that a HIES steering committee is established to provide overall guidance and governance of the survey implementation.

  The steering committee may consist of the head of Statistical office, HIES Manager, HIES Finance officer, HIES technical advisor, representatives from development partners.

• Choice of the Data Capture: paper based, or computer assisted (Month 3, Year 1)

  HIES has been conducted using Computer Assisted Personal Interview (CAPI) method across the Pacific since 2018. This involves using digital tablets, which enhances data quality through built-in validations, consistency checks, and efficient skip patterns. The GPS capability of the tablets aids in field monitoring and provides.

  • Choice of CAPI software

    There are a lot of data collection software and tools that are readily available in the market and some common ones include:

    • World Bank Survey Solutions (SS)
    • Survey123
    • CSPRO Android
    • SurveyCTO

    SPC has recommended the use of Survey solution that is widely used by countries for various data collections including censuses. See this document for CAPI collection using survey solutions.

  • Choice of collection method: recall or diary

    The 7-day recall method has been recommended for HIES in the Pacific, as it can be conducted in one or two visits using CAPI, reducing respondent burden. While diary-based HIES can still be implemented with CAPI, it involves printing the diary for households and manually entering or sending the data to headquarters for processing.

• Questionnaire design (Month 4-6, year 1)

  Designing a questionnaire is crucial for survey preparation, balancing international standards, country needs, and interview length. The HIES questionnaire focuses on collecting poverty data, including food consumption and non-food expenditure. This data helps assess poverty, update Consumer Price Index (CPI) weights, and inform the household component of National Accounts.

  The process also involves translation of questionnaires into other official languages used in the country.

  Sample: Kiribati Household Income and Expenditure Survey 2019

  • The Core HIES Questionnaire for the Pacific

    The core HIES questionnaire is the minimum set of sections and questions that must be collected to report on the main HIES objectives (see introduction): update the basket of goods and services consumed by households for Consumer Price Index update, assess the level of poverty and provide information on a large list of indicators on household living conditions (including SDGs).

  • Additional modules

    The implementation of a HIES survey can be the opportunity to collect additional information, based on country needs or knowledge priorities about households’ or individuals’ characteristics. Additional sections or questions can be added to the core HIES questionnaire upon agreement with the National Statistical Office. They must be directly linked to specific country needs, and the NSO must have identified resources to process and analyse the additional data collected.

  • Questionnaire validations

    The validation is important to ensure relevance of use for the data collected for various purposes such as progress monitoring, reporting obligation and identification of resource gaps. HIES steering committee plays a lead role in the questionnaire design, by conducting the following steps:

    • Questionnaire consultations
    • Questionnaire pretest
    • Questionnaire final approval

• Sampling (4-6 months, year 1)

  The sampling phase is a crucial part of the survey preparation process, as it covers: the sample size computation, and the sample selection (sample distribution). This sampling phase will allow the implementation of the fieldwork plan and the budget.

• Mapping (Month 7, year 1)

  Mapping work occurs at this point of the HIES preparation, and there are two different mapping work that require attention:

  • Harmonisation of EA framework to increase the efficiency of the random selection, it is important that each single EA has similar probability of selection
  • Production of survey maps of the selected enumeration area that highlight all the selected households

• Design Collection (Month 4-6, year 1)

  • HIES field work timeline

    Due to the nature of the data collected (consumption, expenditure, income...), it is important to spread the field work across 12 months within each reporting domains to address seasonality because:

    • Consumption and prices of food items varies according to the season of the year (seasonality of crops, fruit, fishes, vegetables).
    • Some items might not be available during a given month time of the year due to out-of-stock issues and shipment delays.

    The 12-month field operation must consider some holidays for the field staff – usually 4 to 8 weeks, depending on the workload.

  • HIES workload

    The key factor that mainly impacts the HIES workload is the size of the questionnaire, and the time required to complete one interview.

    The interview time is not the only component of the HIES field team workload, and the following tasks must get factored in:

    • Time to access the EA
    • Availability of respondents
    • Time to travel to the next EA for the following round

  • Number of field staff required

    The combination of the sample size, sample distribution the HIES schedule allows to determine the number of field staff required and the breakdown by geographical areas.

  • Training and Field work plan: examples

    Field plans consist of scheduled timeframe for training followed by fieldwork rounds for teams according to their location, and selected sample locations.

• Collection Budgeting and HIES Governance (Month 7-8, year 1)

  The objectives of a HIES budget are:

  • To provide structure and give the project clear guidance regarding its finances.
  • To ensure that proper processes are followed via the standard operating procedure (SOP).
  • To provide regular reporting and updates on the cash flow.
  • To facilitate the allocation of resources by governments and donors.
  • To enable sound financial management and reporting.

  • Government funding

    Here is a generic flow of funding submissions and approvals:

    1. Unit submissions for each division
       • units prepare their budgets
       • verifications
    2. Divisional submission
       • verifications by the division heads
       • verification and justifications by the accounts/ budget team
    3. Department submission
       • verification for clarity by the focal point officer
       • verification by the Ministry’s budget team
    4. Ministry submission
       • Tabled in parliament for final approval

• Donor funding (Month 9-10, year 1)

  Donor funding support usually comes in the form of money and in-kind such as technical assistance and equipment. Once approved by parliament, funding negotiations and lobbying is done through dialoguing with development partners

• Survey Governance (Month 1-36, year 1-3)

  The overall survey governance is to ensure there is flow and continuity in the overall process to minimise any disruptions that might affect the collection and the quality of the data. This including having in place strategies and step by step process for all activities involved.

12 Months – Year 2 to Year 3

The Pacific HIES Toolkit - Fieldwork for HIES

Even though the main field task requires extensive administration with regards to the supervision of enumerators and or revisits, various phases of tasks is also of particular importance which includes:

• Setting up the general administration of the workflow design
• Annual workplan, Operational plan, individual workplan
• Designing of templates and or forms for reporting purposes
• Selection, training and supervision of field workers
• Quality control of the field work
• Arrangements of follow ups for non-response and or revisits for vacant households
• Statistical analysis drawn on sub rounds for data quality checks
• Reporting process whilst in operation

• Recruitment of Field Staff (Month 1-3, year 2)

  National statistics offices (NSOs) oversee the recruitment of field staff, which starts with advertising positions of interviewers and supervisors. The selection process is based on experience in data collection and performance during interviews. In addition to this, it is important to select staff that meet the following:

  • Flexibility in working hours, availability to work after normal working hours, even during weekends
  • Can easily access a vehicle
  • Available to travel in different islands or provinces considering that some trips can last even a few weeks
  • Availability across the 12 months
  • Good communication skills
  • Basic IT knowledge and skills such as navigating through smart devices

• Field staff training (Month 4-6, year 2)

  In preparation for the training the following should be considered:

  • Purchasing of tablet at least 3 months before training
  • Set up of Survey Solutions Server from SPC
  • Training venues, logistics and other equipment need to be organised by the NSO, such as internet access, visual equipment, maps and field manuals

  • Questionnaire and Field Work Operations Manuals

    HIES instructions are usually available in two manuals:

    • The Field work manual, which covers all HIES-related information; and
    • The Questionnaire instruction manual, which details each section and question of the questionnaire, with clear explanations and examples.

  • Training of enumerators – Questionnaire

    Training enumerators on the Survey Questionnaire involves not only field workers but also headquarters and survey management. The training is undertaken for two weeks, with the first week dedicated to HIES and the questionnaire contents and the second week to CAPI, pilot testing and final assessment of the participants

    Based on the training and results of assessments, the enumerators are selected and assigned their roles as either a supervisor or an interviewer. Interviewer reserves are recommended for any turnovers.

• Fieldwork management

  • Interim Data Quality Assessment and Field Worker Refresher Training

    The twelve-month duration of the HIES field work has a significant advantage for the quality of the data collected by field teams: over time, interviewers get more familiar with the questionnaire and some improvements can be expected.

  • Progress monitoring during Operations

    The data and Paradata collected with CAPI systems enable the use of geographic information to ensure a precise monitoring of the teams' location and progress

  • Dealing with delays and other issues

    Delays are expected throughout the fieldwork and having identified earlier helps in minimising the impact of the field work and quality of data

12 Months – Year 2 to Year 3

The Pacific HIES Toolkit - Processing for HIES

Data processing is the process to bring the data from their raw form to data ready for analysis. This is a crucial step to verify numbers and ensure that the figures produced are consistent. The types of data collected make data processing complex because it is susceptible to certain kinds of errors that largely require manual checks such as response errors and outliers. To continue to improve the quality of data, data processing procedures and methods are continuously revised and developed.

• Data Version 1 - Data Preparation and Cleaning (Month 1-3, year 2)

  The version 3 follows stages using data from version 2.

  • Files aggregation and standardisation
  • Creation of the aggregate income and expenditures files
  • Checking the EA codes
  • Final data cleaning and validation
  • Storage of the datasets in version 3

  • Downloading the raw datasets and appending to one dataset.

    The raw datasets are downloaded once the data collection is completed. Datasets can be downloaded from the Survey Solutions server in 3 different formats: Tab separated data, STATA format and SPSS format.

  • Raw data Cleaning and editing

    This step consists in cleaning and editing the raw data. All raw datasets are cleaned and edited using syntaxes developed with the STATA software using the following steps:

    • Clean cover
    • Merge cover and person
    • Check duplicates
    • Merge with household datasets
    • Raw data editing and cleaning
    • Generate tables of basic indicators
    • Creation of Version 1 folder

• Data Version 2 - Item Coding, Data Reshaping and Validation (Month 2, year 2)

  • Recoding

    Additional cleaning and editing are performed on "version 1" datasets by restructuring the files and recoding the items according to international classifications like COICOP, PACCOI, ISCO and ISIC. Each expenditure on non-food and food items is assigned its correct COICOP code according to the description of the items in the survey and the one used in COICOP. The economic activity variables are recoded following ILO standard microdata processing to allow for production of internationally comparable labour market indicators.

    Resources

    • Classification of Individual Consumption According to Purpose (COICOP) 2018
    • International Standard Industrial Classification of All Economic Activities (ISIC), Rev.4
    • International Standard Classification of Occupations (ISCO)

  • Reshaping

    Roster datasets are reshaped wide and merged back to their respective section in the Household or Person datasets in order to have only one observation for each household or person. The datasets are then saved by sections in the folder.

• Data Version 3 - Final Income and Expenditure Data Aggregation

  The version 3 follows stages using data from version 2.

  • Files aggregation and standardisation
  • Creation of the aggregate income and expenditures files
  • Checking the EA codes
  • Final data cleaning and validation
  • Storage of the datasets in version 3

  • Files aggregation and standardisation

    Aggregated files are generated to complete the first step of the data processing. Aggregated files are built for:

    • Cash expenditures (Food and Non-Food)
    • Cash Incomes
    • Gifts received (In Cash and Kind)
    • Home production
    • Imputed rents
    • All the separate files corresponding to each household-level and person-level section of the HIES questionnaire

  • Aggregating all expenditure/income files

    Once the dataset has been thoroughly cleaned and validated, all transaction files related to expenditure and income can be aggregated. SPC constructs an aggregated expenditure dataset by aggregating all the data files listed below into one:

    • Cash expenditure
    • Home production
    • Gifts
    • Barter/exchange
    • Imputed rents
    • Intermediate expenditure

  • Checking the EA codes

    Household locations recorded with the CAPI system can be used to validate the data collected in the field by comparing the GPS coordinates with the EA boundaries, to ensure that the codes entered by the enumerator match with the real location of the dwelling

  • Validation and cleaning

    This data cleaning occurs at the final stage of the process before the release of Version 3. It consists in checking the overall consistency of the data at the household level