Citizenship processing times by country of origin?

[747-captain]

Can you explain how people on this site are a biased sample? In what way do you think the users here deviate from the average person? Are they more likely to have non-routine applications? That is an easy fix as the criteria for an application being non-routine are pretty straight forward. Are they more likely to be more educated? What is the bias?

If I'm not mistaken, I believe what spyfy is trying to say is not that people who frequent this forum possess some sort of inherent traits that will skew the results, but rather that the SAMPLE SIZE that you would get from this forum compared to the number of people who actually apply for citizenship would make the data analysis meaningless.

Any basic probability theory course will tell you that the larger the sample size, the more representative and accurate the results of an analysis will be. For instance, if you get information on 50 people in this forum, and there are 10,000 people applying for citizenship, that data you obtained is simply anecdotal evidence, and does not necessarily point to any trends. You cannot glean any useful information from it!

 

[spyfy]

If I'm not mistaken, I believe what spyfy is trying to say is not that people who frequent this forum possess some sort of inherent traits that will skew the results, but rather that the SAMPLE SIZE that you would get from this forum compared to the number of people who actually apply for citizenship would make the data analysis meaningless.

Any basic probability theory course will tell you that the larger the sample size, the more representative and accurate the results of an analysis will be. For instance, if you get information on 50 people in this forum, and there are 10,000 people applying for citizenship, that data you obtained is simply anecdotal evidence, and does not necessarily point to any trends. You cannot glean any useful information from it!

Yes, that is the most important point.

Also, there is one important selection bias in this forum: People who are facing issues/delays/problems during their application are more likely to visit this forum and/or post here. So there is a lot of bias towards people with applications who are not just being waved through. Simply because if your application never faces issues, you might not even look for a forum like this.

But the more important one is what captain-747 pointed out above: The sample size is way to small. If you don't design your sample appropriately (and you can't do that unless you have a decent background in survey design and empirical methods), 50 out of thousands of applications is just nothing.

So the numbers you'll get from this will be meaningless.

 

[spyfy]

That is easily fixed by looking at regions (this is my first on the spot try, maybe you could divide them a bit further
1. Europe (EU + UK + Norway + Switzerland)
2. Russia, Central Asia plus rest of Europe (from Tirana to Moscow)
3. Middle East and North Africa
4. Sub-Saharan Africa
5. Indian sub-continent (Pakistan, India, Sri Lanka, Bangladesh, Nepal, Bhutan)
6. Latin America
7. China
8. Rest of Asia (Japan, Korea, Indonesia, etc.)

Wait so you want to know how country of origin is linked to application processing time and before collecting any data you somehow categorize countries? Why is geographic location an appropriate way to categorize countries here? Why don't you categorize them by Human Development Index? Or by population? Or by number of immigrants to Canada from that country? Or by visa policy? Or by average age? Or by...

Pre-categorizing the countries along an arbitrary category (here: geography) will just make things worse.

You also cannot seriously cluster a sample of less than 100 people (and you won't get more on this forum).

Do you know how to calculate correlations? p-values? variance? etc?

If you don't, there is no point in trying to survey this data. There is a reason why social science students spend several courses just on survey design and statistics. It's not that easy to get meaningful stats. It's not just "I'll take some info and juggle them around".

Here are some great examples: If you don't do it right, the US spending on science, space and technology correlates with suicides by hanging, strangulation and suffocation:
http://www.tylervigen.com/spurious-correlations

The problem is exactly that there is no justifiable reason to pick exactly those categories of suicide and those categories of spending.

You have to do stats right or its pointless.

Also, even if you find a correlation:
https://en.wikipedia.org/wiki/Correlation_does_not_imply_causation

 

[Jesuslovesyou]

Wait so you want to know how country of origin is linked to application processing time and before collecting any data you somehow categorize countries? Why is geographic location an appropriate way to categorize countries here? Why don't you categorize them by Human Development Index? Or by population? Or by number of immigrants to Canada from that country? Or by visa policy? Or by average age? Or by...

Pre-categorizing the countries along an arbitrary category (here: geography) will just make things worse.

You also cannot seriously cluster a sample of less than 100 people (and you won't get more on this forum).

Do you know how to calculate correlations? p-values? variance? etc?

If you don't, there is no point in trying to survey this data. There is a reason why social science students spend several courses just on survey design and statistics. It's not that easy to get meaningful stats. It's not just "I'll take some info and juggle them around".

Here are some great examples: If you don't do it right, the US spending on science, space and technology correlates with suicides by hanging, strangulation and suffocation:
http://www.tylervigen.com/spurious-correlations

The problem is exactly that there is no justifiable reason to pick exactly those categories of suicide and those categories of spending.

You have to do stats right or its pointless.

In as much as I disagree with the Sample Size/technique, I would still be interested in the findings ( as per Geography). The results are flawed we all know, with lower confidence level and high margin of error but I'm still interested in findings.

 

[walktheline]

In as much as I disagree with the Sample Size/technique, I would still be interested in the findings ( as per Geography). The results are flawed we all know, with lower confidence level and high margin of error but I'm still interested in findings.

This may give you some clues.

http://www.cic.gc.ca/opendata-donneesouvertes/data/IRCC_citz_0001_E.xls

 

[Joshua1]

A lot of issues people raise on these forums are really meant to "run the clock" while they wait to apply or to receive their visas, PR, citizenship etc. These provide entertainment value for everyone. I think the issue of security may come into play when it comes place of birth and "certain" countries lived in visited or worked in for long periods after becoming PR. So, that may affect processing time.

 

[ottawa-pr]

This may give you some clues.

http://www.cic.gc.ca/opendata-donneesouvertes/data/IRCC_citz_0001_E.xls

Thanks! This is fascinating, first of all we had 147,753 new Canadian citizens in 2016 a 41% drop. Second and this I would have never guessed, the number one country of origin is Philippines with nearly 24,000 (that is 16%) followed by India (16,615 or 11%) and China (10,793 or 7%).

 

[ottawa-pr]

You have to do stats right or its pointless.

I don't get why there is so much resistance to a proposed idea. By your standards, nobody should post anything here, the same way you argue about stats you could talk about legal matters and so on and so forth.

I am not trying to publish a peer reviewed academic paper I am trying to figure out if there is a visible difference or not in the sample we have relatively easy access to. Suppose we do this and it turns out people from Middle East take three times as long people from Europe (I think this is very unlikely to be the case but let say it is), then it is by no means conclusive but it means if someone has the time and is willing to put in the effort they can ask questions, talk to reporters, talk to their MP, ask IRCC to release more information, etc.

 

[spyfy]

I don't get why there is so much resistance to a proposed idea. By your standards, nobody should post anything here, the same way you argue about stats you could talk about legal matters and so on and so forth.

I am not trying to publish a peer reviewed academic paper I am trying to figure out if there is a visible difference or not in the sample we have relatively easy access to. Suppose we do this and it turns out people from Middle East take three times as long people from Europe (I think this is very unlikely to be the case but let say it is), then it is by no means conclusive but it means if someone has the time and is willing to put in the effort they can ask questions, talk to reporters, talk to their MP, ask IRCC to release more information, etc.

Nah, do what you want. I'm just trying to explain that what you get will be meaningless.

Also, why are asking legal questions on this forum and trying to make up stats the same thing? You can post and ask about many things that totally make sense. Most threads are like that. I'm just saying the stats will be meaningless.

And do you seriously think a reporter, an MP or IRCC will listen when you say "so I made a survey and some people in an online forum responded. I got 30 replies. And I found some correllation. Oh also I had no way to verify the respondents ID or the truthfulness of their replies." They'll just ignore it. Believe me, they themselves have the stats by country of origin or by whatever other category. They don't need us to get them.

But you do you. I tried to explain why it's pointless, but I'll just keep away from this and let you guys do whatever you want

 

[paulvoyer]

There is only a very small percentage of people who apply for citizenship and post in this forum. If you started trying to collect data you would have insane levels of sampling bias. So you wouldn't get any useful data out of it.

very logic

 

[B0kkie]

Again, those breakdowns would be meaningless. It tells you nothing if you have a sample of fifty people and then see that people from country A need longer to be processed. It would be completely random that you see something in that data.

Imagine you have one guy from country A and it takes 8 months
And you have one guy from country B and it takes 4 months

Would you now claim that country A generally moves slower? No, of course not. And "even" if you have fifty people, selection biased in this forum, that wouldn't change the judgement.

Exactly. And even if you do find that applications for people from certain countries take longer than those from other countries, what does that tell you? Is it discrimination, or is it that getting the information that relates to citizens from one country is more difficult to obtain than information relating to citizens from another country? There are so may variables involved in the process that even if you have some raw data, there will simply be too much missing information to come to any valid analysis.