|Title: <<< Compilation of the House Price Index in Cyprus >>>
The Statistical Service of Cyprus has applied the Rolling Window Hedonic Regression model for the compilation of the Cypriot House Price Index. This method was selected in order to quality-adjust for the characteristics of the dwellings. The methodological work for the revision of the compilation method has been combined with improvements in the methods of data collection in order to increase the accuracy of the index. This paper describes the peculiarities of the Cyprus residential market leading to the specific form of the model. An important conclusion is that a "luxury-house" market with foreign buyers constitutes a significant part of the Cyprus Residential Market. Additionally, the improvements of the data collection methods, leading to data of better quality, are presented. The main contribution of the new collection methods is that additional characteristics of the dwellings can be used. The method of compilation of the House Price Index is explained. The stratification scheme of new and existing dwellings, the explanatory variables introduced in the model in order to quality-adjust for the characteristics of the dwellings and the weighting scheme of sub-indices are described. The regression coe fficients, the coe fficient of determination R2 and the plots of residuals, presented for contracts of sales for a specific quarter, show a good fit for the model. Finally, the paper discusses areas for future improvements on the method of compilation of House Price Index.
|Title: <<< The importance of cognitive interviewing in quality official statistics: The case of smoking habits >>>
Providers of official statistics increasingly recognize a critical role that qualitative pre-field testing has in identifying problematic survey questions. Pre-field methods include cognitive interviews that focus on investigating mental processes involved in answering questions. At the Slovenian National Institute of Public Health, this in-depth qualitative method was employed to thoroughly examine a question on smoking habits, followed by a quantitative testing of the original and revised question in the pilot study. The analysis of cognitive interviewing data revealed the following problems: (I) it was unclear whether all smoking-related products or just tobacco products should be considered, (II) several occasional smokers did not regard themselves as current smokers but rather as former smokers or non-smokers, (III) respondents that have smoked only for a short period in their past regarded themselves as non-smokers or former smokers. Based on these findings, the question on smoking habits was revised: a note was added explaining which products should be considered and two additional answer categories were included, covering occasional smokers and non-smokers that only tried smoking few times. In next step, a pilot study was conducted using computer assisted web interviewing (CAWI) with a simple random sample (SRS) of 600 inhabitants splitted into two halves (the final samples consisted of 82 and 75 respondents), allowing to test and compare the original and revised version of the question. The findings revealed a greater proportion of current smokers (regular and occasional) and a lower proportion of former smokers in a sample answering to the revised question. The proportion of non-smokers (those who never smoked or only tried smoking few times) was largely similar in both samples. Altogether, these findings confirm the importance of cognitive interviewing in identifying problems regarding understanding and answering survey questions, particularly those concerning sensitive topics, prior to collecting data in the field.
|Title: <<< Are we on the same page? Ensuring the synchronisation of data across multiple platforms >>>
It is important for agencies using official data to ensure the data displayed on their websites and in current publications match those published in official source databases. Ensuring such aligns with the principle of accessibility and clarity and is necessary for institutions holding a commitment to data quality. Data transmission infrastructure, including transmission standards and data taxonomies facilitate the transmission of data and subsequently synchronisation of data published by different organisations. However, unobserved obstacles in the data dissemination, transmission, and production processes may occur and hinder the synchronisation. Institutions relying on out of date or incorrect data face damage to their credibility. Even if policy makers use the correct official figures in making decisions, it is important from a public perception point of view that the data published on their website match official figures. We demonstrate two tools to facilitate checking the synchronisation of data. The first compares the data published on the Statistical Data Warehouse of the European Central Bank (ECB) with those officially published by Eurostat. The second tool automatically reads portable document format (PDF) tables prepared for ECB publications and compares the data presented with the source database.
|Title: <<< Impact of applying quality standards on statistical products >>>
This paper aims to improving the quality of statistical products in Egypt through to examine the impact of applying the total quality standards on household survey, focusing on "the income, expenditure and consumption survey" as one of the most important field survey carried out by statistical agencies worldwide, true to the life of the Egyptian citizen. The study used descriptive analytical and quantitative statistical methods, based on the evaluation of the statistical status of Egypt, the European code of best practices, evaluation of experts and citizens on the quality of statistical data, the use of the statistical capacity index as a good means of comparing the performance of the Egyptian statistical apparatus with its counterparts in the Arab, African and international countries. In Egypt, in addition to the SWOT analysis. Statistical methods were used to identify the correlation between the various variables. The study showed that the standard (accuracy) is the most affect by 38% correlated with the other quality standards; then the (availability) is 21%; that both the standard (accuracy) and (availability) affect 59% on the quality of the output of the statistical product, It was also found that 99% of the sample frame design affects the quality of the statistical product, 61% of the response rates in the field work are due to the accuracy and clarity of the statistical form used in the data collection form, only 7% of the researcher's training on the data collection form affects the fieldwork method, it was found that 55% of the analysis of the survey results is due to the accuracy of published data.
|Grazia Di Bella|
|Title: <<< The Istat platform prototype for the documentation of administrative data used as input of the official statistics >>>
Assessing the usability of administrative data (AD), measuring their relevance for the statistics production process, monitoring the AD collection, document the AD treatment are the main objectives of the QRCA (Quality Report Card for AD) internal web site that Istat is going to implement. Many Process and Structural metadata are available directly from the IT tools supporting the AD collection and treatment processes already in use: a) the ARCAM system that provides an interface for the AD owners and a tool to manage collection (150 Administrative sources) and guarantees the security of the data transmission process; b) SIM (System of Integrated Microdata), the DB (Database) of integrated administrative microdata (70 Administrative sources); c) the National Statistical Programme DB reporting statistical uses and making AD accessible for statistical purposes in in terms of compliance with the data privacy law. Once the interoperability of the systems is ensured,these metadata, together with some additional reference metadata, will make availableto internal users to support their statistical production processes. The quality framework used to report (derived from the international project BLUE-ETS), is composed of three Hyperdimensions: Source, Metadata and Data. In each Hyperdimension, dimensions and quality indicators are considered. In the Source Hyperdimension the quality aspects related to the source as a whole are reported. The Metadata Hyperdimension considers conceptual aspects such as the administrative objects (i.e. events and units) and variables and the Data Hyperdimension related to the data quality (facts) includes quality indicators applied to data. For the realization of the prototype the tool that is being used is the Business Intelligence Visual Analytics Platform - MicroStrategy adopted by Istat to realize some data visualization products. Supported by a Java application and a passing Oracle Database, it allows to produce tables and graphs and to generate various types of output.
|Jose Jabier Zurikarai|
|Title: <<< Bringing statistics to users and users to statistics >>>
Statistics are not always as useful, accessible and easy as we would like them to be. The gap between Official Statistics and our users is wide and should be tightened. There are high lever users which can have a fluent relation with us, but what about the rest of them? Students, journalists, local authorities, municipalities and govern department employees and others are not always very familiar with official statistics; sometimes due to their short experience and other times because they are random users. World has gained an increasing complexity for everyone, so we should be aware of the real situation of users. In this context, the achievement of quality framework principles, such as accessibility and clarity, needs some real action. We have to meet the users, talk to them, ask for their expectations, explain the value of our data, how to use data, teach them to be efficient and so on. We believe real interaction has to take place and there are two options: they come to us, or we to go to them. Indeed we are doing and planning both ways. The aim of the paper is to share our experience and future projects. We regularly go to the universities, to the classroom, and teach them what official statistics are and how to use them. We also do similar activities with journalists and public employees. But this is time consuming, so we plan to develop e-learning programs to have large audiences. Activities involving direct interaction are also very valuable because we get some feedback from those students who will be our future high-end users. We will explain what we do and how we do, but also the problems found and the solutions in the making, suitable activities for any agency.