Acquisition of a central administrative register
In some administrative or mixed surveys, data are provided by a single institution. In this situation, no reporting units are used, unless the survey is mixed, i.e. administrative registers are integrated with direct survey data on persons, businesses or institutions.
Ad hoc modul
Statistical production process gathering data by means of a specific section of the questionnaire of the main survey, regularly or occasionally administered to the reporting units or to subset analysis units. Generally, it is aimed at satisfying information needs changing over time.
Administrative act
Act ruling the administrative data formation.
Administrative register
Register containing data on acts, events or statements recorded by public or private institutions by virtue of their institutional role. It can also be used to produce statistical information or to draw and contact units to be surveyed.
Administrative survey
Administrative surveys consist in the acquisition and processing of non-statistical data owned by bodies other than the one in charge of the process, by virtue of their institutional activity, as for example, data collected for management purposes. In administrative surveys, data can be either acquired from several institutions, or single bodies as owner of a central register. In the first case the survey reporting units are the institutions providing data, in the second case a central administrative register provided by a single body is used.
Analysis unit
Major entity observed in the survey. Analysis units can be directly observable sets, e.g. households, household components, enterprises, employees, or sets of events, e.g. holidays, in-patient admissions. Analysis units can coincide with reporting units, when providing information on themselves.
Area
The statistical processes approved in the National Statistical Programme (PSN) are classified according to observed domains, hierarchically broken down into areas, sectors and topics. Each area represents a grouping of sectors.
BANFF/GEIS
BANFF/GEIS are two generalised software programmes for editing and imputation of numerical, continuous and non-negative variables. They have been developed by Statistics Canada and work on different environments. GEIS has been implemented on UNIX; it uses Oracle databases and C executable procedures. BANFF has been implemented on SAS environment and is structured according to SAS philosofy. The two programmes do not differ with respect to the methodology, that is based on conistency rules expressed on linear form.
Buildings Workers' Welfare Funds
Buildings Workers' Welfare Funds (Casse Edili) is a body founded by the building workers' trade unions and the federation of employers of construction industry.
Circular
The circular is a written communication sent within an organisation (enterprise, public administration, ...) to provide orders, regulations or transmit information. Usually is a letter (circular letter) however nowadays it can be an electronic communication (e.g., an e-mail). From the juridical point of view orders and regulations stated in the circular apply to the internal organisation system and, therefore, do not apply to external subjects dealing with it.
Clean data file
It is the data file generated, as output, from the editing and imputation procedure.
Complete production cycle
Set of operations starting from data collection or acquisition of data from input sources and ending with data or estimates dissemination.
CONCORD
CONCORD is a generalised software for data editing and imputation, deriving from the integration of the following softwares: SCIA, GRANADA, RIDA. They were developed and are used by Istat and each of them implements different methods.
Consistency error
Occurrence of the values of two or more data items which do not satisfy some predefined relationship between those data items (Unece).
Court of Accounts
This body was set up under the Constitution (Articles 100 and 103) and is responsible for checking the legality of the activities of the Government and the Public Administration. It is also responsible for checking the management of the accounts of the State, of public administrations generally and of bodies funded by the State. It also has a judicial function in the field of public accounting and in relation to pensions of State officials.
Decree requiring Parliament's approval
Il decreto legislativo consiste in una legge del parlamento che delega il governo a redigere la disciplina di materie soprattutto tecniche. La delega del parlamento al governo non è in bianco: i margini di azione dell'esecutivo sono limitati nel tempo, nell'oggetto e nel contenuto. Non possono esserci quindi deleghe generiche ed inoltre devono essere indicati i principi cui il governo deve attenersi.
Decree with the force of law
Con decreto legislativo o decreto delegato si intende un atto normativo, a disposizione del Governo, avente forza di legge adottato in casi straordinari di necessità ed urgenza, dal potere esecutivo per delega espressa e formale del potere legislativo. Il decreto legge, prima di diventare efficace, deve essere emanato con decreto dal presidente della Repubblica. Il decreto legge dura al massimo sessanta giorni dalla data di emanazione: in questo lasso di tempo si comporta come se fosse una legge, tuttavia l'ultima parola spetta al Parlamento. Se in questi sessanta giorni le Camere non convertono il decreto legge in legge, esso decade e gli effetti prodotti si annullano, come se il decreto legge non fosse mai esistito.
Deterministic imputation
A deterministic imputation method determines one unique value for the imputation of a missing or inconsistent datum. This means that when the imputation process is repeated, the same value is imputed.
Direct survey
In direct or non-administrative surveys, the institution in charge collects and processes data from enterprises, institutions and natural persons. Surveys in which data are collected from public administrations are not necessarily administrative. For instance, the "Survey on social actions and services of single and associated municipalites" consists in the collection of data on services provided by Municipalities and is not based on administrative data already available; it is therefore a direct survey. Direct surveys always use reporting units.
Editing
Data editing is the application of checks that identify missing, invalid or inconsistent entries or that point to data records that are potentially in error. (Statistics Canada Quality Guidelines, 3rd edition, October 1998, page 35) (Eurostat, Doc. Eurostat/A4/Quality/03/Glossary).
Editing and imputation
Check "editing" and "imputation".
Estimation statistics
Data are produced following an estimation methodology. The estimates are obtained on the basis of evaluations by subject matter experts that provide a judgement summarising all available information.
European regulations, directives and decisions
The European Union wields its authority by means of regulations, directives, decisions, advices, reccomendations. The most relevant ones, as sources of our system, are regulations and directives. The regulations take effect immediately, being adressed to the citizenships of the Union Countries, whereas the directive leave free the single state to choose the solutions for their implementation
File standard
Microdata sample sets drawn from some Istat surveys that can be released for study and reseach purposes upon request and authorisation by Istat President. Data are anonimized and any reference permitting the identification of single natural or legal persons is removed (art. 10, Law Decree n. 322/89).
Frame
The materials or devices which delimit, identify, and allow access to the elements of the target population. (Lessler, 1982)
Generalised software
A generalised software for the statistical production is a system aimed to guarantee the production functions in a statistical survey; It has the following caractheristics: i) it implements solid methodologies and statistical techniques; ii) it can be used without or with a limited need of further software development; iii) is equipped with proper documentation and friendly interface, makeing it possible the use also users not skilled from an IT point of view. For an overview of the main softwares used by Istat see: http://www.istat.it/strumenti/metodi/software/.
GRANADA (Rules management and data analysis)
GRANADA performs the imputation of qualitative and quantitative variables according to the deterministic approach, i.e. applying IF (error condition)-THEN (correction action) rules. This module allows to perform the data check (data split into wrong and correct) according to some coherence rules based on logic and arithmetic operators (valid for qualitative and quantitative variables).
Guida Monaci Spa
A datawarehouse in the fields of private enterprises by sector, public administration, health, education, finance, tourism and media.
Imputation
Imputation is the process used to resolve problems of
missing, invalid or inconsistent responses identified during
editing. This is done by changing some of the responses or
missing values on the record being edited to ensure that a
plausible, internally coherent record is created.
(Statistics Canada Quality Guidelines, 3rd edition, October
1998, page 38) (Eurostat, Doc.Imputation is the process used to resolve problems of
missing, invalid or inconsistent responses identified during
editing. This is done by changing some of the responses or
missing values on the record being edited to ensure that a
plausible, internally coherent record is created.
(Statistics Canada Quality Guidelines, 3rd edition, October
1998, page 38) (Eurostat, Doc.Eurostat/A4/Quality/03/Glossary)
Indata
Indata is the Istat website for data capturing.
Input sources of the secondary study
Input sources for secondary studies are documented by specifying the used primary surveys and secondary studies and institutions in charge.
Input-output files
Several phases of the statistical business process imply the creation and the management of intemediate files, generally of data. In SIQual these files are named input-output files, since they represent the ouput of a phase and the input of the following. So far, not all the intermediate products are documented in SIQual, but only the three types: row data, clean data, validated microdata.
Intertel
Company for the planning and production of telecommunication devices.
Italian National Road Administration (Anas)
The Italian National Road Administration is in charge of the national roads and highways. It is a jointstock company owned 100% by the Ministry of Economy and Finance and supervised by the Ministry for Infrastructure and Transport
Law (ordinary)
The formal ordinary law is the result of a juridical procedure involving some bodies: the law has to be approved with the same formulation by the Deputy Chamber and the Republic Senate and promulgated by the Republic President. After his promulgation, the law is published in the Gazzetta Ufficiale and becomes effective in fifteen days from its publication (unless differently stated).
Legislation
Type of formal act determining the realisation of an Istat process.
Ministerial decree and Prime Minister decree
A ministerial decree (DM) is an administrative act issued by a minister on the jurisdiction of the ministry; it is not a law and it may have the form of a secondary law, when it sets regulations. When such an act is issued by the Prime Minister it is referred as decree of the Prime Minister (DPCM).
Missing values
Missing values stem from questions the respondent did not answer. This can happen for several reasons: the respondent may not know the answer, may not be willing to respond or may have simply missed a question.
Mixed survey
Mixed surveys are based both on direct data collection from enterprises, institutions and natural persons and on the acquisition of non statistical data owned by institutions other than the one in charge, by virtue of their institutional activity.
Model-based imputation
Imputation based on an explicitly described statistical model, e.g. use of averages, medians, regression equations, etc. to impute a value.
National Research Council (Italian CNR)
Public body in charge of carrying out, promoting, disseminating, transferring and improving research activities in the main sectors of knowledge and their applications for scientific, technological, economic and social development of the Country.
National Social Security Institute (Italian INPS)
Major Italian social security body.
National Statistical Programme
The National Statistical Programme: a) establishes the statistical processes - and their objectives - of public interest to be carried out by the National Statistical System; b) lasts three years and is updated yearly; c) is set up by Istat (art. 13 legislative decree 322/1989 and following integrations).
Operation
Specific activity aimed at manipulating the information content of the statistical process, implemented by operators and/or by IT procedures. A statistical production process can be split in a set of operations carried out during different phases.
Oversampling
Process consisting in the increasing of the sample size of a survey in a given geographic area, which can correspond to a municipality, a province or a region, with the aim to satisfy specific information needs.
Panel survey
A longitudinal survey in which variables are measured on the same units over time. Panel surveys are particularly useful to understand changes at the individual level.
Periodicity
A statistical process is organised into several phases, each of them may have a single or several periodicities. In general, all phases have the same periodicity called periodicity of the production cycle i.e. the time interval between two successive complete production cycles. If some phases (e.g. data collection, release) have different periodicities, the periodicity of the production cycle is referred to the processing phase, i.e. to the reduced cycle. The data collection periodicity in a primary survey, corresponds to the frequency at which information is requested to investigated units. The estimate periodicity in primary surveys and secondary studies relates to the time reference of the estimates or macrodata released to the users. The data dissemination periodicity in primary surveys and secondary studies, corresponds to the frequency at which official data are released to users (public, Eurostat, ect.). Since process results may be disseminated by different modes, each with different periodicity, the most relevant publication for users is the time reference for the dissemination periodicity. For example, the Multipurpose household survey on travels, holidays and daily life, produces quarterly estimates (periodicity of the estimates) released annually (dissemination periodicity).
Phase
Set of operations interrelated from a temporal-logic point of view. A production process can be split into a sequence of phases that can be carried out in chronological sequence or synchronically.
Phenomenon
Aspects of the real world observed through statistical processes, for instance "natural population balance", "employment", etc..
Pilot survey
A survey, usually on a small scale, carried out prior to the main survey, primarily to gain information to improve the efficiency of the main survey. For example, it may be used to test a questionnaire, to ascertain the time taken by field procedure or to determine the most effective size of sampling unit (OECD Glossary of Statistical Terms).
Postel
Hybrid Mail Service Provider. Poste Italiane company
Primary sampling unit
In multi-stage sampling the sample of elements is obtained as a result of two or more stages of sampling: the population elements are first grouped into disjoint subpopulations, called primary sampling units (PSUs) and a probability sample of PSUs is drawn (first stage sampling). For each PSU in the first-stage sample a probability sample of second-stage sampling units is drawn and so on till the elementary sampling units (Adapted Carl-Erik Sarndal, Bengt Swensson, Jan Wretman (2003) Model Assisted Survey Sampling, Springer Verlag)
Primary survey
Process by which the body in charge produces statistical information through data collection from enterprises, institutions and natural persons and their processing, or through acquisition and processing of non-statistical data owned by bodies other than the body in charge, by virtue of their institutional activity. Currently, Istat is the body in charge of all processes documented in the system. The definition of primary survey includes the following types: direct surveys, surveys from administrative sources, and mixed surveys.
Probabilistic imputation
An imputation rule that is in part a function of a randomization process exogenous to the experimental observations. (Unece)
Process status
It states whether a primary survey or secondary study is underway (active), temporarily suspended or definitively ceased. A process can cease because it is included in or substituted by other processes. A process can be suspended for one or more production cycles as it happens for some surveys when censuses are carried out.
Processing indices
The methodology used to calculate index numbers is described.
Proxy respondent
A proxy is anyone other than the person about whom the information is being collected (Survey methodology, 2004, Groves R.M et al.)
Quality control action
Specific action aimed at preventing, correcting or evaluating nonsampling errors that may arise during the various phases and operations of statistical processes.
Quality control survey or experiment
A replicated survey carried out on a small scale by very experienced staff in order to obtain some “zero-default” results with which the actual results of the survey can be compared (OECD Glossary of Statistical Terms). Sometimes, quality controls are performed by means of experiments which, for Istat classification purposes, are grouped with the quality control surveys.
Questionnaire
Paper or electronic tool used to collect information on analysis units. It is also called survey form. Besides questions and blanks to answer, it may include interviewer instructions, the introduction and visual cards supporting the respondent.
Raw data file
Generally, this file is obtained from the data collection phase, including the controls applied during the interviews, if foreseen, and after the application of coding and manual revision, if planned. It is the data file that used as input of the editing and imputation software.
Reduced production cycle
Set of operations related to revision, pre-processing and processing of data or estimates.
Reference date
Date allowing to shortly describe (format dd-mm-yyyy) the period which data or estimates refer to. For example, for the 2001 General Population Census the reference date is 21-10-2001. Otherwise, when process data do not refer to a day but to time intervals (week, month, year), the last day of the reference period is conventionally used. The reference date also identifies a production cycle.
Reporting unit
Units contacted to obtain information on analysis units. A reporting unit can itself represent an analysis unit or be used to collect information on other units. Reporting units are set for all direct primary surveys and administrative surveys where data are collected from several institutions: for example, health care establishments represent the reporting units for the "Quick survey on hospital discharges"
Response reconciliation
In a re-interview successive to a survey, the reconciliation consists in the ascertainment of the true answer, when the answers to the same question in the two interviews are different.
RIDA (Information reconstruction with automatic donors)
RIDA performs the imputation of qualitative and quantitative variables according to minimum distance donor. Propedeutical activities are: i) classifying units as wrong or correct; ii) recording them into two separate datasets; iii) identifying values to be imputed by means of a given error item.
SCIA (System for the automatic control and imputation)
SCIA implements editing and imputation of qualitative variables strictly following the Fellegi-Holt methodology. For each wrong record the system first identifies the minimum number of variables to be imputed and then performs imputation according to a pre-defined set of check rules.
Secondary sampling unit
Secondary (or second-stage) sampling units are groups of units of the target population that are drawn in the second stage of a multi-stage sampling design within the primary sampling units drawn in the first stage.
Secondary study
Process aimed at producing statistical information, consisting in the processing of statistical data already available from previous primary surveys or secondary studies, or non-statistical data owned by the body in charge, by virtue of its institutional activity.
Sector
The statistical processes approved in the National Statistical Programme (PSN) are classified according to observed domains, hierarchically broken down into areas, sectors and topics. Each sector represents a grouping of topics.
SIGIF (Management system for household surveys)
Sigif is a system used to record the information from different phases of the statistical production process relative to household surveys. It mainly permits to: define the sampling municipalities, draw the sampling households, assign them to the interviewers, transmit the survey material and manage the outcomes concerning nonresponse and anomalies in the data collection phase. The system is integrated to the IVR system (Interactive Voice Response) that allows the interviewers to promptly transmit the results of the contact phase and to the information system municipality collection net (SIRC) that is the database on interviewers.
Sistan
Sistan stands for National Statistical System. It represents the network of Italian private or public bodies providing official statistical information.
Sistat
System for the management of territorial nomenclatures.
Statistical information system
Information system on statistical processes producing: a) data and/or metadata; b) data obtained from several statistical sources totally or partially controlled by the institution in charge of the system; c) integrated or harmonised data with respect to definitions anInformation system on statistical processes producing: a) data and/or metadata; b) data obtained from several statistical sources totally or partially controlled by the institution in charge of the system; c) integrated or harmonised data with respect to definitions and classifications used and/or relevant processes; d) data available to external users at different aggregation levels and disseminated through different channels and modes, for: i) presentation and/or navigation, ii) further processing.
Statistical process
In SIQual, it is a process aiming at producing statistical information, i.e. primary surveys or secondary studies, or statistical information systems.
Statistical register
Database on a target population obtained through validation of a single administrative register, integration of several registers or resulting from an ad hoc survey. It aims at: - ensuring the coverage of the reference universe of relevant statistical units; - identifying units and conditions of their contactability over the territory; - allowing "demographic" analyses of the units by regular recording of relevant events; - drawing lists to carry out statistical surveys.
Subset analysis unit
Units hierarchically dependent on a reporting or an analysis unit. For example, "person aged 18 or over" is a subset unit of "person".
Substitution
The substitution of nonrespondent units is a practice that may only be done using matched pairs that are selected as part of the initial sample selection (National Center for Education Statistics, Statistical Standards).
Summary analysis unit
Units observed as sets of events, e.g. "number of books read".
Survey design
It states whether it is a total survey (as population censuses) or a sample survey, i.e. it observes only a subset of target units and the adopted sampling design is documented.
Target population
Set of units to be studied (Survey methodology, 2004, Groves R.M et al.).
Territorial classification
Most detailed geographical level of dissemination of the results.
Territorial classification
Most detailed level of a national or international standard classification used to disseminate the results.
Thesaurus
List of items used to describe in an analytical and standard way a particular feature of the process, for instance the type of operation.
Time series
Sequence of data points, measured typically at successive times, spaced at (often uniform) time intervals.
Topic
The statistical processes approved in the National Statistical Programme (PSN) are classified according to observed domains, hierarchically broken down into areas, sectors and topics. A topic is the most detailed domain
Validated microdata
It is the file of microdata generated after the validation phase, that may be internal to the process or may use external sources. Therefore is the microdata file allowing the reporducibility of the data published by Istat. Some survey procedures might not allow to distinguish between this file typology and the clean files, thus turing out to have the same content.
Validity period
It identifies the editions in which metadata (survey design, operation, etc.) are used in the process. For Planning and Re-planning operations and their quality control actions the validity period conventionally coincides with the edition in which innovations are introduced.