Sample data for exercises in Further Adventures in Data Cleaning.

Dataset

This dataset was created by Mohanad Hazem Qabil

Contents

Dirty Cafe Sales Dataset

Overview

The Dirty Cafe Sales dataset contains 10,000 rows of synthetic data representing sales transactions in a cafe. This dataset is intentionally "dirty," with missing values, inconsistent data, and errors introduced to provide a realistic scenario for data cleaning and exploratory data analysis (EDA). It can be used to practice cleaning techniques, data wrangling, and feature engineering.

File Information

• File Name: dirty_cafe_sales.csv
• Number of Rows: 10,000
• Number of Columns: 8

Columns Description

Data Characteristics

1. Missing Values:

   • Some columns (e.g., Item, Payment Method, Location) may contain missing values represented as None or empty cells.

2. Invalid Values:

   • Some rows contain invalid entries like "ERROR" or "UNKNOWN" to simulate real-world data issues.

3. Price Consistency:

   • Prices for menu items are consistent but may have missing or incorrect values introduced.

Menu Items

The dataset includes the following menu items with their respective price ranges:

Use Cases

This dataset is suitable for: - Practicing data cleaning techniques such as handling missing values, removing duplicates, and correcting invalid entries. - Exploring EDA techniques like visualizations and summary statistics. - Performing feature engineering for machine learning workflows.

Cleaning Steps Suggestions

To clean this dataset, consider the following steps: 1. Handle Missing Values: - Fill missing numeric values with the median or mean. - Replace missing categorical values with the mode or "Unknown."

1. Handle Invalid Values:

   • Replace invalid entries like "ERROR" and "UNKNOWN" with NaN or appropriate values.

2. Date Consistency:

   • Ensure all dates are in a consistent format.
   • Fill missing dates with plausible values based on nearby records.

3. Feature Engineering:

   • Create new columns, such as Day of the Week or Transaction Month, for further analysis.

License

This dataset is released under the CC BY-SA 4.0 License. You are free to use, share, and adapt it, provided you give appropriate credit.

Feedback

If you have any questions or feedback, feel free to reach out through the dataset's discussion board on Kaggle.

Access and clean an open source herbarium dataset using Excel or RStudio.

Dirty Retail Store Sales Dataset

Overview

The Dirty Retail Store Sales dataset contains 12,575 rows of synthetic data representing sales transactions from a retail store. The dataset includes eight product categories with 25 items per category, each having static prices. It is designed to simulate real-world sales data, including intentional "dirtiness" such as missing or inconsistent values. This dataset is suitable for practicing data cleaning, exploratory data analysis (EDA), and feature engineering.

File Information

• File Name: retail_store_sales.csv
• Number of Rows: 12,575
• Number of Columns: 11

Columns Description

Categories and Items

The dataset includes the following categories, each containing 25 items with corresponding codes, names, and static prices:

Electric Household Essentials

Furniture

The dissertation_demo.zip contains the base code and demonstration purpose for the dissertation: A Conceptual Model for Transparent, Reusable, and Collaborative Data Cleaning. Each chapter has a demo folder for demonstrating provenance queries or tools. The Airbnb dataset for demonstration and simulation is not included in this demo but is available to access directly from the reference website. Any updates on demonstration and examples can be found online at: https://github.com/nikolausn/dissertation_demo

Restaurant Menu DatasetWith approximately 45,000 menus dating from the 1840s to the present, The New York Public Library’s restaurant menu collection is one of the largest in the world. The menu data has been transcribed, dish by dish, into this dataset. For more information, please see http://menus.nypl.org/about.This dataset is not clean and contains many missing values, making it perfect to practice data cleaning tools and techniques.Dataset Variables:id: identifier for menuname: sponsor: who sponsored the meal (organizations, people, name of restaurant)event: categoryvenue: type of place (commercial, social, professional)place: where the meal took place (often a geographic location)physical_description: dimension and material description of the menuoccasion: occasion of the meal (holidays, anniversaries, daily)notes: notes by librarians about the original materialcall_number: call number of the menukeywords: language: date: date of the menulocation: organization or business who produced the menulocation_typecurrency: system of money the menu uses (dollars, etc)currency_symbol: symbol for the currency ($, etc)status: completeness of the menu transcription (transcribed, under review, etc)page_count: how many pages the menu hasdish_count: how many dishes the menu has

All data are prone to error and require data cleaning prior to analysis. An important example is longitudinal growth data, for which there are no universally agreed standard methods for identifying and removing implausible values and many existing methods have limitations that restrict their usage across different domains. A decision-making algorithm that modified or deleted growth measurements based on a combination of pre-defined cut-offs and logic rules was designed. Five data cleaning methods for growth were tested with and without the addition of the algorithm and applied to five different longitudinal growth datasets: four uncleaned canine weight or height datasets and one pre-cleaned human weight dataset with randomly simulated errors. Prior to the addition of the algorithm, data cleaning based on non-linear mixed effects models was the most effective in all datasets and had on average a minimum of 26.00% higher sensitivity and 0.12% higher specificity than other methods. Data cleaning methods using the algorithm had improved data preservation and were capable of correcting simulated errors according to the gold standard; returning a value to its original state prior to error simulation. The algorithm improved the performance of all data cleaning methods and increased the average sensitivity and specificity of the non-linear mixed effects model method by 7.68% and 0.42% respectively. Using non-linear mixed effects models combined with the algorithm to clean data allows individual growth trajectories to vary from the population by using repeated longitudinal measurements, identifies consecutive errors or those within the first data entry, avoids the requirement for a minimum number of data entries, preserves data where possible by correcting errors rather than deleting them and removes duplications intelligently. This algorithm is broadly applicable to data cleaning anthropometric data in different mammalian species and could be adapted for use in a range of other domains.

Ahoy, data enthusiasts! Join us for a hands-on workshop where you will hoist your sails and navigate through the Statistics Canada website, uncovering hidden treasures in the form of data tables. With the wind at your back, you’ll master the art of downloading these invaluable Stats Can datasets while braving the occasional squall of data cleaning challenges using Excel with your trusty captains Vivek and Lucia at the helm.

In this project, we work on repairing three datasets:

• Trials design: This dataset was obtained from the European Union Drug Regulating Authorities Clinical Trials Database (EudraCT) register and the ground truth was created from external registries. In the dataset, multiple countries, identified by the attribute country_protocol_code, conduct the same clinical trials which is identified by eudract_number. Each clinical trial has a title that can help find informative details about the design of the trial.
• Trials population: This dataset delineates the demographic origins of participants in clinical trials primarily conducted across European countries. This dataset include structured attributes indicating whether the trial pertains to a specific gender, age group or healthy volunteers. Each of these categories is labeled as (`1') or (`0') respectively denoting whether it is included in the trials or not. It is important to note that the population category should remain consistent across all countries conducting the same clinical trial identified by an eudract_number. The ground truth samples in the dataset were established by aligning information about the trial populations provided by external registries, specifically the CT.gov database and the German Trials database. Additionally, the dataset comprises other unstructured attributes that categorize the inclusion criteria for trial participants such as inclusion.
• Allergens: This dataset contains information about products and their allergens. The data was collected from the German version of the `Alnatura' (Access date: 24 November, 2020), a free database of food products from around the world `Open Food Facts', and the websites: `Migipedia', 'Piccantino', and `Das Ist Drin'. There may be overlapping products across these websites. Each product in the dataset is identified by a unique code. Samples with the same code represent the same product but are extracted from a differentb source. The allergens are indicated by (‘2’) if present, or (‘1’) if there are traces of it, and (‘0’) if it is absent in a product. The dataset also includes information on ingredients in the products. Overall, the dataset comprises categorical structured data describing the presence, trace, or absence of specific allergens, and unstructured text describing ingredients.

N.B: Each '.zip' file contains a set of 5 '.csv' files which are part of the afro-mentioned datasets:

• "{dataset_name}_train.csv": samples used for the ML-model training. (e.g "allergens_train.csv")
• "{dataset_name}_test.csv": samples used to test the the ML-model performance. (e.g "allergens_test.csv")
• "{dataset_name}_golden_standard.csv": samples represent the ground truth of the test samples. (e.g "allergens_golden_standard.csv")
• "{dataset_name}_parker_train.csv": samples repaired using Parker Engine used for the ML-model training. (e.g "allergens_parker_train.csv")
• "{dataset_name}_parker_train.csv": samples repaired using Parker Engine used to test the the ML-model performance. (e.g "allergens_parker_test.csv")

Data Cleansing Tools Market size was valued at USD 4.02 Billion in 2024 and is projected to reach USD 9.20 Billion by 2032, growing at a CAGR of 10.89% during the forecast period 2026-2032.Demand for Accurate Data Analytics: A strong demand for accurate datasets is being noticed, and the use of data cleansing techniques is expected to expand to enable trustworthy reporting and decision-making.Adoption of Cloud Platforms: Enterprise workloads are being moved to the cloud, and cloud-compatible data cleansing solutions are expected to be used to boost scalability and flexibility.

This dataset is designed specifically for beginners and intermediate learners to practice data cleaning techniques using Python and Pandas.

It includes 500 rows of simulated employee data with intentional errors such as:

Missing values in Age and Salary

Typos in email addresses (@gamil.com)

Inconsistent city name casing (e.g., lahore, Karachi)

Extra spaces in department names (e.g., " HR ")

✅ Skills You Can Practice:

Detecting and handling missing data

String cleaning and formatting

Removing duplicates

Validating email formats

Standardizing categorical data

You can use this dataset to build your own data cleaning notebook, or use it in interviews, assessments, and tutorials.

Dataset

This dataset was created by Martin Kanju

Released under Other (specified in description)

Contents

Data cleaning is one of the most important but time-consuming tasks for data scientists. The data cleaning task consists of two major steps: (1) error detection and (2) error correction. The goal of error detection is to identify wrong data values. The goal of error correction is to fix these wrong values. Data cleaning is a challenging task due to the trade-off among correctness, completeness, and automation. In fact, detecting/correcting all data errors accurately without any user involvement is not possible for every dataset. We propose a novel data cleaning approach that detects/corrects data errors with a novel two-step task formulation. The intuition is that, by collecting a set of base error detectors/correctors that can independently mark/fix data errors, we can learn to combine them into a final set of data errors/corrections using a few informative user labels. First, each base error detector/corrector generates an initial set of potential data errors/corrections. Then, the approach ensembles the output of these base error detectors/correctors into one final set of data errors/corrections in a semi-supervised manner. In fact, the approach iteratively asks the user to annotate a tuple, i.e., marking/fixing a few data errors. The approach learns to generalize the user-provided error detection/correction examples to the rest of the dataset, accordingly. Our novel two-step formulation of the error detection/correction task has four benefits. First, the approach is configuration free and does not need any user-provided rules or parameters. In fact, the approach considers the base error detectors/correctors as black-box algorithms that are not necessarily correct or complete. Second, the approach is effective in the error detection/correction task as its first and second steps maximize recall and precision, respectively. Third, the approach also minimizes human involvement as it samples the most informative tuples of the dataset for user labeling. Fourth, the task formulation of our approach allows us to leverage previous data cleaning efforts to optimize the current data cleaning task. We design an end-to-end data cleaning pipeline according to this approach that takes a dirty dataset as input and outputs a cleaned dataset. Our pipeline leverages user feedback, a set of data cleaning algorithms, and a set of previously cleaned datasets, if available. Internally, our pipeline consists of an error detection system (named Raha), an error correction system (named Baran), and a transfer learning engine. As our extensive experiments show, our data cleaning systems are effective and efficient, and involve the user minimally. Raha and Baran significantly outperform existing data cleaning approaches in terms of effectiveness and human involvement on multiple well-known datasets.

Netflix is a popular streaming service that offers a vast catalog of movies, TV shows, and original contents. This dataset is a cleaned version of the original version which can be found here. The data consist of contents added to Netflix from 2008 to 2021. The oldest content is as old as 1925 and the newest as 2021. This dataset will be cleaned with PostgreSQL and visualized with Tableau. The purpose of this dataset is to test my data cleaning and visualization skills. The cleaned data can be found below and the Tableau dashboard can be found here .

Data Cleaning

We are going to: 1. Treat the Nulls 2. Treat the duplicates 3. Populate missing rows 4. Drop unneeded columns 5. Split columns Extra steps and more explanation on the process will be explained through the code comments

--View dataset

SELECT * 
FROM netflix;

--The show_id column is the unique id for the dataset, therefore we are going to check for duplicates
                                  
SELECT show_id, COUNT(*)                                                                                      
FROM netflix 
GROUP BY show_id                                                                                              
ORDER BY show_id DESC;

--No duplicates

--Check null values across columns

SELECT COUNT(*) FILTER (WHERE show_id IS NULL) AS showid_nulls,
    COUNT(*) FILTER (WHERE type IS NULL) AS type_nulls,
    COUNT(*) FILTER (WHERE title IS NULL) AS title_nulls,
    COUNT(*) FILTER (WHERE director IS NULL) AS director_nulls,
    COUNT(*) FILTER (WHERE movie_cast IS NULL) AS movie_cast_nulls,
    COUNT(*) FILTER (WHERE country IS NULL) AS country_nulls,
    COUNT(*) FILTER (WHERE date_added IS NULL) AS date_addes_nulls,
    COUNT(*) FILTER (WHERE release_year IS NULL) AS release_year_nulls,
    COUNT(*) FILTER (WHERE rating IS NULL) AS rating_nulls,
    COUNT(*) FILTER (WHERE duration IS NULL) AS duration_nulls,
    COUNT(*) FILTER (WHERE listed_in IS NULL) AS listed_in_nulls,
    COUNT(*) FILTER (WHERE description IS NULL) AS description_nulls
FROM netflix;

We can see that there are NULLS. 
director_nulls = 2634
movie_cast_nulls = 825
country_nulls = 831
date_added_nulls = 10
rating_nulls = 4
duration_nulls = 3 

The director column nulls is about 30% of the whole column, therefore I will not delete them. I will rather find another column to populate it. To populate the director column, we want to find out if there is relationship between movie_cast column and director column

-- Below, we find out if some directors are likely to work with particular cast

WITH cte AS
(
SELECT title, CONCAT(director, '---', movie_cast) AS director_cast 
FROM netflix
)

SELECT director_cast, COUNT(*) AS count
FROM cte
GROUP BY director_cast
HAVING COUNT(*) > 1
ORDER BY COUNT(*) DESC;

With this, we can now populate NULL rows in directors 
using their record with movie_cast 

UPDATE netflix 
SET director = 'Alastair Fothergill'
WHERE movie_cast = 'David Attenborough'
AND director IS NULL ;

--Repeat this step to populate the rest of the director nulls
--Populate the rest of the NULL in director as "Not Given"

UPDATE netflix 
SET director = 'Not Given'
WHERE director IS NULL;

--When I was doing this, I found a less complex and faster way to populate a column which I will use next

Just like the director column, I will not delete the nulls in country. Since the country column is related to director and movie, we are going to populate the country column with the director column

--Populate the country using the director column

SELECT COALESCE(nt.country,nt2.country) 
FROM netflix AS nt
JOIN netflix AS nt2 
ON nt.director = nt2.director 
AND nt.show_id <> nt2.show_id
WHERE nt.country IS NULL;
UPDATE netflix
SET country = nt2.country
FROM netflix AS nt2
WHERE netflix.director = nt2.director and netflix.show_id <> nt2.show_id 
AND netflix.country IS NULL;


--To confirm if there are still directors linked to country that refuse to update

SELECT director, country, date_added
FROM netflix
WHERE country IS NULL;

--Populate the rest of the NULL in director as "Not Given"

UPDATE netflix 
SET country = 'Not Given'
WHERE country IS NULL;

The date_added rows nulls is just 10 out of over 8000 rows, deleting them cannot affect our analysis or visualization

--Show date_added nulls

SELECT show_id, date_added
FROM netflix_clean
WHERE date_added IS NULL;

--DELETE nulls

DELETE F...

Autonomous Data Cleaning with AI Market Outlook

According to our latest research, the global Autonomous Data Cleaning with AI market size reached USD 1.68 billion in 2024, with a robust year-on-year growth driven by the surge in enterprise data volumes and the mounting demand for high-quality, actionable insights. The market is projected to expand at a CAGR of 24.2% from 2025 to 2033, which will take the overall market value to approximately USD 13.1 billion by 2033. This rapid growth is fueled by the increasing adoption of artificial intelligence (AI) and machine learning (ML) technologies across industries, aiming to automate and optimize the data cleaning process for improved operational efficiency and decision-making.

The primary growth driver for the Autonomous Data Cleaning with AI market is the exponential increase in data generation across various industries such as BFSI, healthcare, retail, and manufacturing. Organizations are grappling with massive amounts of structured and unstructured data, much of which is riddled with inconsistencies, duplicates, and inaccuracies. Manual data cleaning is both time-consuming and error-prone, leading businesses to seek automated AI-driven solutions that can intelligently detect, correct, and prevent data quality issues. The integration of AI not only accelerates the data cleaning process but also ensures higher accuracy, enabling organizations to leverage clean, reliable data for analytics, compliance, and digital transformation initiatives. This, in turn, translates into enhanced business agility and competitive advantage.

Another significant factor propelling the market is the increasing regulatory scrutiny and compliance requirements in sectors such as banking, healthcare, and government. Regulations such as GDPR, HIPAA, and others mandate strict data governance and quality standards. Autonomous Data Cleaning with AI solutions help organizations maintain compliance by ensuring data integrity, traceability, and auditability. Additionally, the evolution of cloud computing and the proliferation of big data analytics platforms have made it easier for organizations of all sizes to deploy and scale AI-powered data cleaning tools. These advancements are making autonomous data cleaning more accessible, cost-effective, and scalable, further driving market adoption.

The growing emphasis on digital transformation and real-time decision-making is also a crucial growth factor for the Autonomous Data Cleaning with AI market. As enterprises increasingly rely on analytics, machine learning, and artificial intelligence for business insights, the quality of input data becomes paramount. Automated, AI-driven data cleaning solutions enable organizations to process, cleanse, and prepare data in real-time, ensuring that downstream analytics and AI models are fed with high-quality inputs. This not only improves the accuracy of business predictions but also reduces the time-to-insight, helping organizations stay ahead in highly competitive markets.

From a regional perspective, North America currently dominates the Autonomous Data Cleaning with AI market, accounting for the largest share in 2024, followed closely by Europe and Asia Pacific. The presence of leading technology companies, early adopters of AI, and a mature regulatory environment are key factors contributing to North America’s leadership. However, Asia Pacific is expected to witness the highest CAGR over the forecast period, driven by rapid digitalization, expanding IT infrastructure, and increasing investments in AI and data analytics, particularly in countries such as China, India, and Japan. Latin America and the Middle East & Africa are also gradually emerging as promising markets, supported by growing awareness and adoption of AI-driven data management solutions.

Component Analysis

The Autonomous Data Cleaning with AI market is segmented by component into Software and Services. The software segment currently holds the largest market share, driven

This dataset includes all of the data downloaded from GBIF (DOIs provided in README.md as well as below, downloaded Feb 2021) as well as data downloaded from SCAN. This dataset has 2,808,432 records and can be used as a reference to the verbatim data before it underwent the cleaning process. The only modifications made to this datset after direct download from the data portals are the following:

1) for GBIF records, I renamed the countryCode column to be "country" so that the column title is consistent across both GBIF and SCAN 2) A source column was added where I specify if the record came from GBIF or SCAN 3) Duplicate records across SCAN and GBIF were removed by identifying identical instances "catalogNumber" and "institutionCode" 4) Only the Darwin core columns (DwC) that were shared across downloaded datasets were retained. GBIF contained ~249 DwC variables, and SCAN data contained fewer, so this combined dataset only includes the ~80 columns shared between the two datasets

For GBIF, we downloaded the data in three separate chunks, therefore there are three DOIs. See below:

GBIF.org (3 February 2021) GBIF Occurrence Downloadhttps://doi.org/10.15468/dl.6cxfsw GBIF.org (3 February 2021) GBIF Occurrence Downloadhttps://doi.org/10.15468/dl.b9rfa7 GBIF.org (3 February 2021) GBIF Occurrence Downloadhttps://doi.org/10.15468/dl.w2nndm

Quadrant provides Insightful, accurate, and reliable mobile location data.

Our privacy-first mobile location data unveils hidden patterns and opportunities, provides actionable insights, and fuels data-driven decision-making at the world's biggest companies.

These companies rely on our privacy-first Mobile Location and Points-of-Interest Data to unveil hidden patterns and opportunities, provide actionable insights, and fuel data-driven decision-making. They build better AI models, uncover business insights, and enable location-based services using our robust and reliable real-world data.

We conduct stringent evaluations on data providers to ensure authenticity and quality. Our proprietary algorithms detect, and cleanse corrupted and duplicated data points – allowing you to leverage our datasets rapidly with minimal processing or cleaning. During the ingestion process, our proprietary Data Filtering Algorithms remove events based on a number of both qualitative factors, as well as latency and other integrity variables to provide more efficient data delivery. The deduplicating algorithm focuses on a combination of four important attributes: Device ID, Latitude, Longitude, and Timestamp. This algorithm scours our data and identifies rows that contain the same combination of these four attributes. Post-identification, it retains a single copy and eliminates duplicate values to ensure our customers only receive complete and unique datasets.

We actively identify overlapping values at the provider level to determine the value each offers. Our data science team has developed a sophisticated overlap analysis model that helps us maintain a high-quality data feed by qualifying providers based on unique data values rather than volumes alone – measures that provide significant benefit to our end-use partners.

Quadrant mobility data contains all standard attributes such as Device ID, Latitude, Longitude, Timestamp, Horizontal Accuracy, and IP Address, and non-standard attributes such as Geohash and H3. In addition, we have historical data available back through 2022.

Through our in-house data science team, we offer sophisticated technical documentation, location data algorithms, and queries that help data buyers get a head start on their analyses. Our goal is to provide you with data that is “fit for purpose”.

E-commerce Product Dataset - Clean and Enhance Your Data Analysis Skills or Check Out The Cleaned File Below!

This dataset offers a comprehensive collection of product information from an e-commerce store, spread across 20+ CSV files and encompassing over 80,000+ products. It presents a valuable opportunity to test and refine your data cleaning and wrangling skills.

What's Included:

A variety of product categories, including:

• Apparel & Accessories
• Electronics
• Home & Kitchen
• Beauty & Health
• Toys & Games
• Men's Clothes
• Women's Clothes
• Pet Supplies
• Sports & Outdoor
• (and more!)

Each product record contains details such as:

• Product Title
• Category
• Price
• Discount information
• (and other attributes)

Challenges and Opportunities:

Data Cleaning: The dataset is "dirty," containing missing values, inconsistencies in formatting, and potential errors. This provides a chance to practice your data-cleaning techniques such as:

• Identifying and handling missing values
• Standardizing data formats
• Correcting inconsistencies
• Dealing with duplicate entries

Feature Engineering: After cleaning, you can explore opportunities to create new features from the existing data, such as: - Extracting keywords from product titles and descriptions - Deriving price categories - Calculating average discounts

Who can benefit from this dataset?

• Data analysts and scientists looking to practice data cleaning and wrangling skills on a real-world e-commerce dataset
• Machine learning enthusiasts interested in building models for product recommendation, price prediction, or other e-commerce tasks
• Anyone interested in exploring and understanding the structure and organization of product data in an e-commerce setting
• By contributing to this dataset and sharing your cleaning and feature engineering approaches, you can help create a valuable resource for the Kaggle community!

Abstract

The main objectives of the survey were: - To obtain weights for the revision of the Consumer Price Index (CPI) for Funafuti; - To provide information on the nature and distribution of household income, expenditure and food consumption patterns; - To provide data on the household sector's contribution to the National Accounts - To provide information on economic activity of men and women to study gender issues - To undertake some poverty analysis

Geographic coverage

National, including Funafuti and Outer islands

Analysis unit

• Household
• individual

Universe

All the private household are included in the sampling frame. In each household selected, the current resident are surveyed, and people who are usual resident but are currently away (work, health, holydays reasons, or border student for example. If the household had been residing in Tuvalu for less than one year: - but intend to reside more than 12 months => The household is included - do not intend to reside more than 12 months => out of scope

Kind of data

Sample survey data [ssd]

Sampling procedure

It was decided that 33% (one third) sample was sufficient to achieve suitable levels of accuracy for key estimates in the survey. So the sample selection was spread proportionally across all the island except Niulakita as it was considered too small. For selection purposes, each island was treated as a separate stratum and independent samples were selected from each. The strategy used was to list each dwelling on the island by their geographical position and run a systematic skip through the list to achieve the 33% sample. This approach assured that the sample would be spread out across each island as much as possible and thus more representative.

For details please refer to Table 1.1 of the Report.

Sampling deviation

Only the island of Niulakita was not included in the sampling frame, considered too small.

Mode of data collection

Face-to-face [f2f]

Research instrument

There were three main survey forms used to collect data for the survey. Each question are writen in English and translated in Tuvaluan on the same version of the questionnaire. The questionnaires were designed based on the 2004 survey questionnaire.

HOUSEHOLD FORM - composition of the household and demographic profile of each members - dwelling information - dwelling expenditure - transport expenditure - education expenditure - health expenditure - land and property expenditure - household furnishing - home appliances - cultural and social payments - holydays/travel costs - Loans and saving - clothing - other major expenditure items

INDIVIDUAL FORM - health and education - labor force (individu aged 15 and above) - employment activity and income (individu aged 15 and above): wages and salaries, working own business, agriculture and livestock, fishing, income from handicraft, income from gambling, small scale activies, jobs in the last 12 months, other income, childreen income, tobacco and alcohol use, other activities, and seafarer

DIARY (one diary per week, on a 2 weeks period, 2 diaries per household were required) - All kind of expenses - Home production - food and drink (eaten by the household, given away, sold) - Goods taken from own business (consumed, given away) - Monetary gift (given away, received, winning from gambling) - Non monetary gift (given away, received, winning from gambling)

Questionnaire Design Flaws Questionnaire design flaws address any problems with the way questions were worded which will result in an incorrect answer provided by the respondent. Despite every effort to minimize this problem during the design of the respective survey questionnaires and the diaries, problems were still identified during the analysis of the data. Some examples are provided below:

Gifts, Remittances & Donations Collecting information on the following: - the receipt and provision of gifts - the receipt and provision of remittances - the provision of donations to the church, other communities and family occasions is a very difficult task in a HIES. The extent of these activities in Tuvalu is very high, so every effort should be made to address these activities as best as possible. A key problem lies in identifying the best form (questionnaire or diary) for covering such activities. A general rule of thumb for a HIES is that if the activity occurs on a regular basis, and involves the exchange of small monetary amounts or in-kind gifts, the diary is more appropriate. On the other hand, if the activity is less infrequent, and involves larger sums of money, the questionnaire with a recall approach is preferred. It is not always easy to distinguish between the two for the different activities, and as such, both the diary and questionnaire were used to collect this information. Unfortunately it probably wasn?t made clear enough as to what types of transactions were being collected from the different sources, and as such some transactions might have been missed, and others counted twice. The effects of these problems are hopefully minimal overall.

Defining Remittances Because people have different interpretations of what constitutes remittances, the questionnaire needs to be very clear as to how this concept is defined in the survey. Unfortunately this wasn?t explained clearly enough so it was difficult to distinguish between a remittance, which should be of a more regular nature, and a one-off monetary gift which was transferred between two households.

Business Expenses Still Recorded The aim of the survey is to measure "household" expenditure, and as such, any expenditure made by a household for an item or service which was primarily used for a business activity should be excluded. It was not always clear in the questionnaire that this was the case, and as such some business expenses were included. Efforts were made during data cleaning to remove any such business expenses which would impact significantly on survey results.

Purchased goods given away as a gift When a household makes a gift donation of an item it has purchased, this is recorded in section 5 of the diary. Unfortunately it was difficult to know how to treat these items as it was not clear as to whether this item had been recorded already in section 1 of the diary which covers purchases. The decision was made to exclude all information of gifts given which were considered to be purchases, as these items were assumed to have already been recorded already in section 1. Ideally these items should be treated as a purchased gift given away, which in turn is not household consumption expenditure, but this was not possible.

Some key items missed in the Questionnaire Although not a big issue, some key expenditure items were omitted from the questionnaire when it would have been best to collect them via this schedule. A key example being electric fans which many households in Tuvalu own.

Cleaning operations

Consistency of the data: - each questionnaire was checked by the supervisor during and after the collection - before data entry, all the questionnaire were coded - the CSPRo data entry system included inconsistency checks which allow the NSO staff to point some errors and to correct them with imputation estimation from their own knowledge (no time for double entry), 4 data entry operators. - after data entry, outliers were identified in order to check their consistency.

All data entry, including editing, edit checks and queries, was done using CSPro (Census Survey Processing System) with additional data editing and cleaning taking place in Excel.

The staff from the CSD was responsible for undertaking the coding and data entry, with assistance from an additional four temporary staff to help produce results in a more timely manner.

Although enumeration didn't get completed until mid June, the coding and data entry commenced as soon as forms where available from Funafuti, which was towards the end of March. The coding and data entry was then completed around the middle of July.

A visit from an SPC consultant then took place to undertake initial cleaning of the data, primarily addressing missing data items and missing schedules. Once the initial data cleaning was undertaken in CSPro, data was transferred to Excel where it was closely scrutinized to check that all responses were sensible. In the cases where unusual values were identified, original forms were consulted for these households and modifications made to the data if required.

Despite the best efforts being made to clean the data file in preparation for the analysis, no doubt errors will still exist in the data, due to its size and complexity. Having said this, they are not expected to have significant impacts on the survey results.

Under-Reporting and Incorrect Reporting as a result of Poor Field Work Procedures The most crucial stage of any survey activity, whether it be a population census or a survey such as a HIES is the fieldwork. It is crucial for intense checking to take place in the field before survey forms are returned to the office for data processing. Unfortunately, it became evident during the cleaning of the data that fieldwork wasn?t checked as thoroughly as required, and as such some unexpected values appeared in the questionnaires, as well as unusual results appearing in the diaries. Efforts were made to indentify the main issues which would have the greatest impact on final results, and this information was modified using local knowledge, to a more reasonable answer, when required.

Data Entry Errors Data entry errors are always expected, but can be kept to a minimum with