Wednesday, July 26, 2017

Qts on PCCF/PCCF+

Question
I am trying to get a better understanding of the PCCF/PCCF+ products in order to assist its use by health researchers at Dalhousie. I will use an example to illustrate my questions: A researcher is working with patient data for the time period (year) 2008 to 2011 including postal codes. He would like to use the PCCF/PCCF+ product to better identified each patient’s locality. My questions are:


E.g. 1. Census 2006 has multiple PCCF+ version 5 releases.

Release Including Postal Codes through
c Mar 2008
d Sep 2008
e Mar 2009
f Jul 2009
g Dec 2009
h Oct 2010
j May 2011
k May 2011


Should he be using multiple releases of the PCCF+ to correspond with the patient year data? That is PCCF+ 5d for patient data in year 2008, PCCF+ 5g, PCCF+ 5h for patient data in year 2009, PCCF+ 5k for patient data in year 2009? Or can he just use PCCF+ 5k to identity patient locality info for the full 4 years patient data?

E.g. 2. If the patient data only came with FSA, is/are there any advantage(s) in using PCCF+ over PCCF? If not, given that there are multiple PCCF releases (see below), should he be matching up the PCCF releases with the patient year data? Can he just use the latest PCCF release to allocate patient locality info for the full 4 years patient data?

E.g. 1. Census 2006 has multiple PCCF+ version 5 releases.

Release Including Postal Codes through
Pccf_sep2008 Sep 2008
Pccf_dec2009 Dec 2009
Pccf_oct2010 Oct 2010
Pccf_may2011 May 2011


Answer
Thank you for the question. We have received the following response from Subject Matter:

“With regards to your client’s situation, I would suggest using the PCCF+ over the PCCF because the PCCF+ provides correspondence to historical dissemination/enumeration areas (back to 1981) and it also processes partial postal codes, such as FSAs (although the geocoding is less precise than full 6-digit postal code). 

In general, we recommend always using the latest letter version of a PCCF+ census release because they are the cleanest versions of the product.  For example, if your client chose to use the PCCF+ based to the 2006 Census then he would use version 5K.  Having said that, I would recommend that he uses the most recent version of the product, PCCF+ 6D (based to the 2011 Census geography), because it will assign both the 2011 and 2006 DAs to the patient record and it also includes improvements that are not part of version 5. 

FYI, I have attached a document which provides some guidance on when to use the PCCF+ over the PCCF, along with a more detailed overview presentation."

Wednesday, July 19, 2017

Smoking habits of young Canadians

Question
I have a very strong feeling that the answer to this question lies in a custom tabulation, but I figured I’d throw it out in case anyone knows of a resource I missed…

I have a researcher who is looking into the habits of young legal Canadian smokers. Of course, this means they want different age ranges depending on the province of origin (18-20 or 19-20 depending). They were using CCHS 2014, but are open to using another source if they can get the data they need – the closest I could get was using the Youth Smoking Survey, which lets us group the ages how we’d like, but since it’s looking at students and not the whole population in the age range it’s still not quite right. Is there a survey I missed that might let them get the data for the age ranges I mentioned?

Response
The PUMF for the Canadian Tobacco, Alcohol and Drugs Survey (CTAD) might work. The 2015 person file has an "age of respondent" variable (dvage) with detailed ages from 15 to 85. There's a variable for provinces (prov). And the survey certainly includes a lot of variables about smoking! It's more of a question of seeing if the ages can be grouped together with large enough amounts of respondents for your researcher's analysis to make sense.

Bonus content: It just so happens that I use "smoking" as an example search in an Odesi exercise I sometimes give to students (attached). We don't subscribe to Odesi, but it has great metadata and I often use it to identify potential PUMFs of interest on different topics, using pretty much the same procedure as described in the exercise. Searching Nesstar is another great way to go about it. Please note that this last link might not work from off-campus.

Tuesday, July 18, 2017

Canadian Cancer Registry documentation

Question
I’m trying to help a researcher who’s interested in using the CCR. For the moment all he needs is documentation (data dictionary, code book, or even just a variables list) to determine if it’s worth going the RDC route, but the items linked on the CCR page at StatCan are rather dated, the newest having a 2008 release date. Is there newer documentation available?

Answer
Attached is the 2015/2016 edition of the CCR Data Dictionary. The author division did clarify that the most recent documents will be made public and put online once they’re fully translated.

Monday, July 17, 2017

Student responses about culture of higher ed. institutes

Question
I have an inquiry from a researcher who is looking for responses from students about the culture and behaviours found in higher educational institutes. I am new to DLI and as such am sure that I haven’t looked in all the possible places, but if such data exists I would appreciate being pointed in the right direction!

Answer
I couldn’t find anything from a quick search on our website, so I’ve asked our Education team for information. Please stay tuned for a response.

Unfortunately the education team does not have any data collected for student culture and behaviour. Do you mind elaborating on what the researcher is specifically looking for? We might be able to find something similar if we can get a sense of the context of these terms.

Most current SHS PUMF

Question
Is the PUMF for the 2015 SHS available? I’m finding loads of tables, and Nesstar has the 2009 PUMF, but that’s all I’m seeing.

Followup
I last asked this question in March. The SHS was revised and changed in 2010. No PUMF has been done since the revision. We keep asking but the author division has not had the resources to produce a PUMF.

We will keep making the case that even doing a PUMF every couple of years would be a huge benefit from or this very important and heavily used survey.

Answer
Unfortunately the last PUMF produced for this survey was in 2009, but I have passed this thread along to the author division and I’ve asked them whether they have any news on releasing a PUMF. Unfortunately there is no plan yet in the decision to produce a SHS PUMF. The more recent data remains the SHS PUMF 2009. As Peter said, we’ll continue to make the case for a PUMF to be produced and highlight its importance in the data community.

Ultrafiltered milk exports and imports

Question
There is a researcher interested in getting the import and exports in Canada and US for ultrafiltered milk. The CIMT has trade data on milk that is full cream, partially and completely skimmed milk not ultrafiltered milk.

Is US and Canada import/export data available on ultrafiltered milk?


Answer
Unfortunately the division responsible checked and could not find a Harmonized Classification System code specified for ultra-filtered milk, whether for exports or imports.

HS codes are employed to track the international trade of commodities; this specific product isn’t tracked.

Wednesday, July 12, 2017

Most popular Census release

Question
Of the standard census releases below,  which is the most popular ?

Population and dwelling counts - Age and sex - Type of dwelling - Families, households and marital status – Language – Income - Immigration and ethnocultural diversity – Housing - Aboriginal peoples – Education – Labour - Journey to work - Language of work -Mobility and migration

That is, which ones do they  get the most questions about and the most interest on day of release ?  Which gets the most media interest?

I would have thought that Pop. And Dwelling counts gets the most buzz,  just because it is the first release, and people are interested in communities changes in population size. I would have thought that income would be the next most popular, giving a look at poverty changes. But does  the Census division have any information on whether these assumptions are correct or not. They may be able to comment more on past census releases, rather than the current one which is not finished yet. What was most popular in 2006 and 2011 would be great.

Answer
The most popular standard census release is Pop and Dwell by a large margin.

2011 Census release
http://www12.statcan.gc.ca/census-recensement/2016/ref/release-dates-diffusion-eng.cfm

NHS release
http://www.statcan.gc.ca/daily-quotidien/130814/dq130814a-eng.htm

Friday, July 7, 2017

CTUMS variance tables for 2007

Question
Can anyone tell me why the variance tables for CTUM 2005 are available on the Statistics Canada website but 2007 aren't?

2005 seems to be available from http://www23.statcan.gc.ca/imdb/p2SV.pl?Function=getSurvDocument&Item_Id=48423&InstaId=22693&ai=1&dis=1 but all I can find for 2007 are the user guides.

Answer
Variability tables for CTUMS 2007 can also be found on the DLI PUMF EFT safe: /MAD_PUMF_FMGD_DAM/Root/4440_CTUMS_ESUTC/2007/annual/doc

If you require any assistance with accessing your EFT account, please let us know.

I inquired whether they would be placed online and unfortunately there are no plans to put them online; but they are part of our release package for the PUMF that the DLI receives and uploads to the EFT.

Wednesday, July 5, 2017

Offenses by Aboriginal individuals leading to incarceration

Question
Are there any data on this, anywhere? A student is looking for information on the types of crimes committed by people of Aboriginal origin that lead to incarceration of some kind. I've found some reports on the Public Safety Canada website, but the offence categories given are quite broad. What I imagine is something like what appears in CANSIM table 251-0013, but for adults and youth, and where information on the Aboriginal offenders is available.

Answer
We have received the following response from subject matter regarding your request:

"I have looked into the request and the details of the offense and incarceration (like in table 251-0013 CANSIM) is no longer collected, so the available data are until 2014. We will start collecting these data again and will be releasing them in the future, but not at this point in time."

Impact of years to graduate (undergraduate and graduate degrees) on wages

Question
A student is looking for data in order to conduct a regression analysis.

She writes I am “trying to see if taking gap years or just taking longer time to graduate for other reasons will effect the wage of that person.The data that I am looking at should have wages, and some other variable that will be able to tell me if the person took longer time than expected to graduate.”

Data on level of degree obtained is available, but we can’t locate anything on years to complete the degree. Canadian data is preferable, but U.S. would be taken if necessary.

Additional Question
Would number of years taken to complete a graduate degree be available as well? I am looking for Canadian data.

Answer
Subject matter has provided the following suggestion for this request:

“I took a look at the NGS codebook and there is variables that would provide you with the duration of an education program. The variables [that] would give you the start date of the program are:
- PR_Q06A - In what year did you start your program at (your educational institution)?
- PR_Q06B - In what month was that?

Then the variables [that would give the end date of a program are]:
- PR_Q11A - In what year did you complete the requirements for your (certificate / diploma / degree) program? By "complete the requirements" I mean write the last exam, submit the last paper, report or assignment, or successfully defend your thesis.
- PR_Q11B - In what month was that?

If we use these 4 variables we are able to find the duration of the program.”

I was unable to locate any additional surveys apart from the NGS; this would be the best source to start with.

Followup Question
Thankyou but unfortunately, those variables aren’t included in the PUMF. The income variables aren’t either. In this case, the student needs it for a paper so a custom tabulation, even if available, will be out of the question.

However, for the future, are custom tabs available on the NGS?

Followup Answer
Unfortunately those variables are only available in the Masterfile. As for your question, subject matter has confirmed that it is possible to create custom tables for the NGS.

CCHS 2015 - Nutrition Component Availability

Question
Could you please tell me when the CCHS 2015 Nutrition component data will be available in the RDC?

Also, I believe I saw that someone recently asked on the list about the CCHS 2015 PUMF and that the response was that there wasn’t going to be a PUMF released for just the 2015 data or the nutrition data. Am I remembering this correctly?

Answer
The CCHS 2015 Nutrition data has been available in the RDC’s as of Friday June 29th 2017. They are now available for use in the RDCs. Furthermore, the CCHS team has confirmed the following: There is a CCHS Nutrition PUMF in the works. No committed release date as yet. I would say end of 2017 at the earliest, but more likely sometime in 2018. We’ll likely have a better idea on the timing at the end of the summer.

Dropout rates for Universities

Question
Where would I find drop out rates for universities in Canada?

Answer
Unfortunately the education division has confirmed that they do not collect data on drop out rates.

Tuesday, July 4, 2017

DB-level data for Halifax

Question
I have a faculty member preparing a brief for City Council regarding a proposed heritage designation for a small downtown neighbourhood in Halifax, consisting of Dissemination Blocks 12090855009, 12090854004, 12090854003 and 12090965001. He is looking for any data that may show the array of incomes within this small area, which he hypothesizes to be a very wide spread (from poor students to wealthy homeowners).

We have some data of this sort, from the NHS 2011, at the level of Dissemination Areas, but, as always, the smaller DB geography is needed if possible. I’ve also found household income ranges by postal routes, but postal route geography is also larger than the dissemination blocks that constitute this neighbourhood.

Is there any way in which income data (households or individuals, averages, medians, etc.) at the level of these DBs might be made available to a client?

Response
According to the 2016 Census Dictionary: Only population and dwelling counts are disseminated at the dissemination block level (with the dissemination area being the smallest standard geographic area for which characteristic data are disseminated). The Chief from the Tax Section of ISD confirmed that the 2 possible geographic options using tax data would be the DA or postal code.

Furthermore, the Data Quality and Confidentiality Standard and Guidelines for the 2011 NHS states that income data is suppressed when an area has a population less than 250 or the number of private households is less than 40.

I’m currently checking with the regional office to confirm that by combining DB’s together both the population and household requirements are met so they should be able to get the information that you’re looking for.

Canadian labour turnover rates

Question
A student is looking for Canadian labour/employee turnover rates, broken down by industry. Are there any current surveys that would provide this data?

The best source I’ve found so far is the Labour Force Survey PUMF which identifies “Job leavers” in the “Flows into unemployment” variable (FLOWUNEM). Data can be cross-tabulated by broad industry using 21 NAICS codes (NAICS_21).

Am I overlooking a better source? I have seen mentions of other data sources in various Statcan publications:

Longitudinal Employment Analysis Program:  CANSIM tables from this data do not identify job leavers or turnover rate
Workplace and Employment Survey:  included data on turnover, but survey discontinued
Survey of Employment, Payrolls and Hours:  CANSIM tables available don’t seem to include data on turnover

I see that this question has been asked before (see archived answer from 2007 in DIGRS), but I wondered what the best source would be now?

I noticed some discussion of a new Statcan survey funded by ESDC which would include turnover data, but (as of 2014) the findings had not been released. I’m not sure which survey the following sources are referring to, nor whether the data is available now:

Drummond, D. (2014, June). Wanted: Good Canadian labour market information. IRPP Insight, 6. Retrieved from Institute for Research on Public Policy website, http://irpp.org/wp-content/uploads/2014/06/insight-no6.pdf
“ESDC has piloted a workplace survey that focuses on skills shortages, hard-to-fill vacancies, employment and turnover.” (p. 13)

Bradshaw, J., & Grant, T. (2014, April 22) Lack of funds keeps Statscan study from public. Globe and Mail. Retrieved from.
https://www.theglobeandmail.com/report-on-business/economy/lack-of-funds-keeps-statscan-study-from-public/article18088971/


Answer
The Labour Statistics Division has confirmed that the LFS cannot produce turnover rates. Unfortunately the also confirmed that do not know of any survey in Statistics Canada that could produce this.

Manitoba Basic Annual Income Experiment data

Announcement
There has recently been renewed interest, in the news, regarding a universal basic income. This discussion sometimes refers to Mincome or the Manitoba Basic Annual Income Experiment. This experiment took place in Winnipeg and Dauphin back in the 1970s. Some of you might be interested to know that the data and documentation for this can now be accessed at:

Mincome overview and links to documentation
http://gregorymason.ca/mincome/ 

Mincome data - six Excel data files and documentation
http://dataverse.lib.umanitoba.ca/dataverse/Mincome_Data 

Revisiting Manitoba's basic-income experiment / Professor Gregory Mason. 
http://news.umanitoba.ca/op-ed-clarifying-the-impacts-of-the-mincome-experiment/ 

Professor Mason was part of the "Institute for Social and Economic Research", which used to be located at the University of Manitoba. This is the agency that documented and processed the Mincome data. Professor Mason is largely responsible for resurrecting this data.

Credit should also go to the University of Toronto (Leanne Trimble) and the University of British Columbia (Paul Lesack) for supplying us with the raw data and some of the documentation. 

I believe both sites retained this data thanks to the foresight of Laine Ruus.

Finally, credit should also go to Lisa O'Hara, Carell  Jackimiek, and Mullai Manickavalli for setting up the University of Manitoba dataverse, where the data can now be accessed.

Further to...
Wow!  What a wonderful message.  And to add to the good news, Prof. Evelyn Forget, a Health Economist from the University of Manitoba has revisited the data and conducted a quasi-experimental analyses.  If your researchers are interested in the data and her work, you might direct them to her.  She's also part of the CNRDC network of research data centres across Canada.

The Town with No Poverty: The Health Effects of a Canadian Guaranteed Annual Income Field Experiment | Canadian Public Policy