Upload
austen-quinn
View
223
Download
0
Embed Size (px)
Citation preview
Handling Reference Handling Reference QuestionsQuestions
DLI Orientation SessionDLI Orientation Session
Kingston, OntarioKingston, Ontario
April 5, 2004April 5, 2004
How much you do will depend How much you do will depend on the level of service you offer on the level of service you offer
in your data centre!in your data centre!
Worst reference question?Worst reference question?
I need some dataI need some data
Reference interview questionsReference interview questions
Why do you need data?Why do you need data?
What type of data do you need?What type of data do you need?
What are you looking for?What are you looking for?
What geographic area(s) do you need?What geographic area(s) do you need?
What time period do you want?What time period do you want?
Why do you need data?Why do you need data?
Any number of reasons:Any number of reasons: I need data to do stats withI need data to do stats with I need to know that there is data available, but I need to know that there is data available, but
I don’t actually have to use it (yet)I don’t actually have to use it (yet) I need it for my thesisI need it for my thesis I need to prepare a sample for my class to I need to prepare a sample for my class to
use on their exercisesuse on their exercises I need to know if we have it for a grant I need to know if we have it for a grant
proposalproposal
Types of data needs (1)Types of data needs (1)
A data file dealing with a specific topicA data file dealing with a specific topic
A data file that contains specific variablesA data file that contains specific variables
Data files that can be compared (for Data files that can be compared (for countries, time periods, regions)countries, time periods, regions)
Types of data needs (2)Types of data needs (2)
Aggregate dataAggregate data
Time series dataTime series data
MicrodataMicrodata
Geospatial dataGeospatial data
MapMap
Aggregate dataAggregate data
Type of dataType of data
Level of aggregation (geographic)Level of aggregation (geographic)
Level of aggregation (unit of analysis)Level of aggregation (unit of analysis)
FormatFormat
E.g, Exports from Ontario of pig iron in Excel formatE.g, Exports from Ontario of pig iron in Excel format
Time series dataTime series data
Aggregate dataAggregate data
Adds time to an aggregate questionAdds time to an aggregate question
Which stats package is Which stats package is veryvery important important May be able to convert between formatsMay be able to convert between formats
MicrodataMicrodata
Unit of analysis (individual, household, Unit of analysis (individual, household, family, business)family, business)
Level of geographic detail neededLevel of geographic detail needed
Topics Topics (usually more than one)(usually more than one)
E.g., health and income of London individualsE.g., health and income of London individuals
Geospatial DataGeospatial Data
Adds the ability to link data to mapsAdds the ability to link data to maps May want to link microdata or aggregate data May want to link microdata or aggregate data
(e.g., respondents’ location, average income)(e.g., respondents’ location, average income)
Requires establishing level (and format) of Requires establishing level (and format) of geographic linkgeographic link
Add on all previous requirementsAdd on all previous requirements
Geospatial linkingGeospatial linking
May link unrelated data sourcesMay link unrelated data sources
Postal Code Conversion file allows Postal Code Conversion file allows mapping of characteristics of individual mapping of characteristics of individual respondents (by postal code) with census respondents (by postal code) with census socio-demographics of census tractsocio-demographics of census tract
You may have to find out if this is what You may have to find out if this is what your user needs – they may not express ityour user needs – they may not express it
MapsMaps
Format of map Format of map (e.g., Arcview, Mapinfo)(e.g., Arcview, Mapinfo)
Geographic coverage Geographic coverage (e.g, CMA of London, or (e.g, CMA of London, or CSD of city of London)CSD of city of London)
Characteristics of map Characteristics of map (e.g., street networks, (e.g., street networks, waters, rail lines, electrical transmission lines, elevations)waters, rail lines, electrical transmission lines, elevations)
Currency of map Currency of map (e.g., what time is captured)(e.g., what time is captured)
CongratulationsCongratulations
You have found out what the patron wants, You have found out what the patron wants, (or thinks they want) and why(or thinks they want) and why
That That maymay have been the easy part have been the easy part
Now, you have to find it …Now, you have to find it …
Where to lookWhere to look
http://idls.ssc.uwo.ca/idls/presentation/DLIhttp://idls.ssc.uwo.ca/idls/presentation/DLIOrientation/Orientation/referencetools19972004.docreferencetools19972004.doc (see handout): (see handout): tthis document provides various reference sources to his document provides various reference sources to help with locating materialshelp with locating materials
Google Google cancan work – or can deliver garbage work – or can deliver garbage
It works well to find organizations; not so It works well to find organizations; not so well for units within gov’t departmentswell for units within gov’t departments
Be creativeBe creative
Think of alternate terms or approaches for Think of alternate terms or approaches for a topic a topic (e.g., unemployment may be hidden under (e.g., unemployment may be hidden under labour force activity)labour force activity)
Look at the question backwards – who Look at the question backwards – who might have collected the desired data – might have collected the desired data – then look for the organization instead of then look for the organization instead of the data the data (e.g., tobacco use (e.g., tobacco use → → Non-smoker’s rights Non-smoker’s rights organizations)organizations)
Be carefulBe careful
Certain organizations are biasedCertain organizations are biased
Government organizations can be biasedGovernment organizations can be biased
Get data sets from opposed organizations Get data sets from opposed organizations and compare them if no neutral data existand compare them if no neutral data exist
Try to help users evaluate data reliabilityTry to help users evaluate data reliability
Tools for finding variablesTools for finding variables
Statistics Canada Thematic Search ToolStatistics Canada Thematic Search Toolhttp://www.statcan.ca/english/Tst/ssint.htmhttp://www.statcan.ca/english/Tst/ssint.htm
Other networked search capabilitiesOther networked search capabilities IDLSIDLS
http://idls.ssc.uwo.cahttp://idls.ssc.uwo.ca QWIFSQWIFS
http://library.queensu.ca/webdoc/ssdc/access_external.htmlhttp://library.queensu.ca/webdoc/ssdc/access_external.html SherlockSherlock
http://http://sherlock.crepuq.qc.ca/public/anglais/recherche.htmlsherlock.crepuq.qc.ca/public/anglais/recherche.html
What next?What next?
What you do with the patron after you What you do with the patron after you identify the needed data depends on your identify the needed data depends on your level of servicelevel of service
Don’t ignore codebooks as a reference Don’t ignore codebooks as a reference source or as a technical resourcesource or as a technical resource
What you do with them if you can’t provide What you do with them if you can’t provide the data is the flip side of that issue.the data is the flip side of that issue.
Can’t find a data file?Can’t find a data file?
Maybe it doesn’t exist (hasn’t been Maybe it doesn’t exist (hasn’t been collected) – new approach to a subjectcollected) – new approach to a subject
May not be available in electronic format May not be available in electronic format (Civil Aviation…)(Civil Aviation…)
May be non-released administrative dataMay be non-released administrative data
Can’t provide a data file?Can’t provide a data file?
May not be publicly availableMay not be publicly available Proprietary data (esp. business)Proprietary data (esp. business) Confidential dataConfidential data Administrative dataAdministrative data
May not be able to obtain itMay not be able to obtain it Can’t afford to purchase itCan’t afford to purchase it Can’t obtain it (e.g., some education data is Can’t obtain it (e.g., some education data is
not distributed outside of United States)not distributed outside of United States) Data have been lostData have been lost
What else?What else?
Ask questions – DLIlist, SOS-DATA, Ask questions – DLIlist, SOS-DATA, individual contacts, etc.individual contacts, etc.
Refer patrons to others who can help:Refer patrons to others who can help: Remote access, Research Data Centre, Remote access, Research Data Centre,
colleague, StatsCan division, etc.colleague, StatsCan division, etc.
Be lazy!Be lazy!
See if other people have done the work See if other people have done the work and will share (e.g., SPSS syntax)and will share (e.g., SPSS syntax)
Record what you have done – you don’t Record what you have done – you don’t want to have to do the same work again!want to have to do the same work again!
Share the work that you have done – Share the work that you have done – assist othersassist others
Final wordsFinal words
This morning, I said working with data is This morning, I said working with data is fun – reference is what makes it fun and fun – reference is what makes it fun and satisfying (instant or delayed gratification)satisfying (instant or delayed gratification)
Reference can define your data serviceReference can define your data service Providing what you have is easyProviding what you have is easy Letting patrons know it’s there isn’t easyLetting patrons know it’s there isn’t easy Providing what you don’t have is difficultProviding what you don’t have is difficult