ACS FAQ: ID's
Datasets are used in research for analysis. If these datasets or tables contain data about persons, one wants to keep track of what has been collected for each person. This is done by using ID's. In medical research it is good pratice to use anonymous ID's, which means you can not trace a person based on the ID: using dates of birth or initals is not allowed.
An example of an ID is "M10142". If in a trial this person's temperature is taken on februari 23 then there will be a table "Temperature" with a record like "M10142", "23-3-2000", "38".
This person can also participate in an other trial, using an other ID, f.i. "D1104580315". In this trial her weigth is recorded. There will now be a table "Weight" with a record like: "H1104580315", "30-8-1999", "78".
In these two examples "M10142" and "D1104580315" are ID's: in ACS such ID's are called "ID in table", to distinguish them from the so called ACS ID's. These examples show that one person can have two different ID's in tables.
ACS ID's are different from ID's in tables in that there are no datasets that use ACS ID's. ACS ID's are used to combine, for one person, data collected in different datasets. This is done by recording in the ACS database that the person, known under ACS ID "B003" has records in tables "Temperature" and in "Weight": and also that in "Temperature" she is known as "M10142" and in table "Weight" als "D1104580315".
There are two ways of looking at this question: you may want to know what has been collected for a person and you know the ACS ID of this person. You choose the form "find a ACS ID" and you type in the ID: a list is displayed with the tables that have records for this person, plus the ID that is used in these tables and the number of records.
It is also possible that all you know about this person is that she is known in a certain dataset as "Q23985". Choose the form "find an ID in table" and type in "Q23985": a list is given of tables that have one or more records with "Q23985" as ID. You can now choose to see a list of all tables that have records for this person and in which she is known as "Q23985", or a list of all tables that have records for her, regardsless of the ID's that were used.
No, you can not view the data. All you can see are the names and descriptions of the tables and the fields, plus the number of records per person in a table, but you can not view the actual data.
For each dataset that is stored on the ACS server there is a so-called DSN file. DSN is for DataSourceName: with the information in this file it is possible to access the data using ODBC (OpenDataBaseConnectivity). If it is known for a table which field has the ID, the table can be queried to see how many and what ID's are used and how many records there are for each ID.
Using the same technique is also possible to get a list of all tables, fields and field descriptions.
ID groups are sets of tables. In all these tables, each person has exactly one ID and exactly one ACS ID. This is done to make sure that f.i. all tables used for the analysis of a trial use consistent ID's. Or to assure all the ID's for an institution are used correctly.
Choose errors in ID groups to see a list of all ID's that are in conflict with these two rules.
You may want to combine two or more tables. You can use the cross table check to see how many persons these tables have in common and also how much records there are for these persons.
When you want the combine the data from two different tables and the ID's used in these tables are not the same, you can use the make link-table utility. With it you choose the two tables you want to combine. A small Access database is made with two tables in it: one contains just two records with information about the two tables. The other table has two rows: for each person the ID's in the two tables are stored in these fields. You can download the database and use this link-table to combine the data in the two tables.