STUCK @ NVC for AP

[cdneh 3,112]

Wow..talk about extensive!!! That is excellent work! I was pretty sure there was something new going on there at NVC...

[Talyss 0]

That's a lot of hard work. Thanks for making the effort and sharing it with us

[bryb 0]

All right gang, well I finally got around to compiling some stats on our NVC wait. I'd been meaning to do this for a few weeks.

To show you just how PATHETIC is the life of a guy waiting for his honey's visa so they can finally be reunited ..... yes, I actually am spending my Valentines night doing this!

How's that for a guy with literally no life???

Some technical details, for those so inclined...

First of all - there is A LOT of data in the Visajourney database. As we all knew by looking at Igor's page, it is possible to get great stuff from this database. Igor has obviously mastered the ability to run a cron-based job to automatically pull the dataset hourly. He seems to run SQL queries directly against the VJ database, Perhaps he's figured out the VJ http syntax to format his queries via the webserver but it doesn't look to me as though that's what he's doing.

Anyway, I don't have direct SQL access to the db and I'm not smart enough to decipher the http parameter strings. So I just hacked it manually.

Instead I just used the built in query builder on the VJ website itself. That kind of sucks because it's pretty weak, I can't run the queries I'd really want to. It lets you hide columns, but the row filtering was not helpful. I just grabbed all rows in the database for which "NVC Received" was non-null. That was 7113 rows out of a total 16,310 in the database. After playing around to narrow my data range as best I could, I just grabbed the resulting html pages manually and cut/pasted the raw data into Excel.

I then spent hours tonight working with Excel DCOUNTA and DAVERAGE functions to get the stats I wanted. Excel is garbage, imho. I should have done it in all in Perl, it would have taken me far less time. I don't know what possessed me to do it in Excel. Oh well.

Unfortunately VJ does not have a field on our timeline data for "AP at NVC". So it's not possible to know explicitly which applications got stuck in NVC AP.

What our timelines do have is the Date Entered NVC and Date Left NVC. Even then, unfortunately, not all VJ members recorded both of these fields. In many cases they record only one or the other or neither, so I can't compute the time spent at NVC.

So what I did was compute the time spent at NVC for those who recorded both values.

The earliest record with 'NVC received' date in the database is 8/23/1995 That's got to be a typo. The 'NVC left' date for that record was in 2008. What, they spent 13 YEARS in NVC??? Um, I doubt it. So ok, that prompted me to toss all records for which NVC Left-NVC entered was too big (I picked 1000 days, that threw away all the junk like this one).

I then compared the time spent at NVC to a threshold value which I felt was large enough to indicate that any application which spent that long at NVC probably got stuck in AP. I picked the threshold at 15 days.

That's important to remember!!! If you pick a different threshold (10 days, 20 days), you'll get different results! Caveat emptor.

The data generally covers the period 2003-2009, which seems to be the period VJ has been operational. There are some data from < 2003, but very few, and I suspect they are typos. As you'll see I focused most of my attention on the past year, 2008.

Anyway, based upon a 15 day threshold I then computed break-down stats for K1 and K3 filings, and I did that month-by-month for 2008 to see if things trended worse at the end of last year. Guess what - no surprise to all of us stuck at NVC, they ARE trending worse!!!

OK, here are the results:

Total number of rows in VJ database: 16,310

Number of rows that have a 'Start NVC' date: 7,113 (=== this is the total number of raw data rows imported into Excel spreadsheet)

Number of rows that are K1s : 5,002

Number of rows that are K3s: 836

Number of rows that are IR1/CR1: 1,275

Well, that's not really so interesting. We really care about rows that have BOTH start and end NVC dates:

Number of records that have start/end NVC dates 4,696

Number of records that don't have start/end NVC dates 2,417

Number of K1s that have start/end NVC dates 3,520

Number of K3s that have start/end NVC dates 603

Number of IR1/CR1s that have start/end NVC dates 573

OK, now here things start to get interesting. These are the average times (in days) for each visa category spent at NVC over the entire database history (2003-2009):

Average time in NVC for all records 20.349

Average time in NVC for K1s 6.918

Average time in NVC for K3s 10.121

Average time in NVC for IR1/CR1s 113.621

As we always knew, K1's generally get in and out quickly - an average of 6.9 days. K3s are a bit longer (10 days), and IR1/CR1 stay MUCH longer because that's where most of their processing occurs.

Right. Next.

Here is the K1 data, broken down month by month since Jan 08:

               Num K1/month    Avg NVC time    Num K1 > threshold    Avg NVC time > threshold    % K1 in month > threshold
Jan-08    111                    6.414            7                            35.4286                            6%
Feb-08    126                    5.651            7                            26.2857                           6%
Mar-08    125                    4.744            5                            21.2000                           4%
Apr-08    56                    4.714            4                            19.7500                           7%
May-08    78                    4.795            2                            28.5000                           3%
Jun-08    46                    4.761            1                            32.0000                           2%
Jul-08    95                    4.137            4                            19.0000                           4%
Aug-08    58                    6.448            4                            31.0000                           7%
Sep-08    71                    5.577            6                            24.3333                           8%
Oct-08    69                    4.710            3                            26.6667                           4%
Nov-08    64                    10.125            11                            43.2727                           17%
Dec-08    73                    12.452            15                            44.0000                           21%
Jan-09    116                    3.845            0                            #DIV/0!                           0%
Feb-09    13                    2.231            0                            #DIV/0!                           0%

Here's how to interpret this. Look at the first row for Jan 08, for example. In that month there were 111 K1 visas that entered NVC and also exited at some point. Those 111 applications spent an average of 6.4 days at NVC. Seven of those visas spent 15 days or longer at NVC, and they each spent an average of 35.4 days at NVC - good chance that they were 'AP' cases. 7/111 is 6%, so that's approximately the percentage of all Jan08 K1s which got stuck in AP.

Ignore the Jan/Feb 09, since those contain records which have entered NVC but not yet exited > 15 days ... That's US FOLKS!!!! STUCK NOW AT NVC!!! Hence those rows have 'zeroes' for Num K1 > threshold, and #Div/0 for the avg time > threshold.

OK, so now if you look down through the months we see something disturbing. In Nov and Dec 2008, BOTH of the last two columns jump up.

Suddenly in Nov/Dec the average time that the 'long' cases (those > 15 days which we suspect are AP) are spending an average of 43 or 44 days at NVC which is considerably higher than previous months. And it gets close to our current anecdotal evidence on this VJ thread that the wait is about 50 days.

Also, we see in Nov/Dec that 17% and 21% , respectively, of all K1 cases are waiting more than 15 days and so are suspect of being AP.

That tells us two things:::::

There seem to be MORE AP cases happening. And when they happen, they are stuck in AP longer.

Sigh.

Well, at least now we know it's not our imagination. It's really happening.

I did a similar analysis for the K3's but since there's much less data, the statistics are less relevant (law of large numbers etc.)

I've attached 2 charts showing this end of year trend in the K1 data.

I tried to attach my Excel file with all the data - all raw data, calculations, and the charts of the K1 2008 data. However VJ does not seem to allow me to update .xls files.

If you would like to see this file, just PM me and I will email it to you.

Happy Valentines all.

Looking forward to getting out of NVC HELL

Thanks for this information great work.

[Badgerella 5]

Wow!

Great work!

[scandal 951]

OK, here is some more analysis.

I did several more things.

First, I went back to all the data for the past two years - 2007 and 2008.

I also played around with the threshold, looking at NVC times of > 10 days, 15 days, 20 days, 40 days.

In all cases, the trend of a spike up in Nov, Dec 2008 is apparent. However it is most dramatic when looking at > 40 days, so that's what I'm presenting here.

It turns out there are 36 K-1 applications over the past 2 years 2007 and 2008 which took > 40 days at NVC

These 36 are the "stuck in NVC AP" cases.

22 of these 36 (61%) occurred in the 2 months of Nov/Dec 2008.

Also, there is duplication and I believe suspect data in the earliest records from 2007. Four of those records are reported from China and I believe are duplicates. Tossing those means 22/33 = 67% are in the last 2 months. In any event it's clear SOMETHING is going on at NVC recently.

Q: What is the average wait time in NVC for those who get stuck?

A: The average wait time for all 36 is 62 days. However the suspect records from early 2007 had wait times > 100 days. If we don't count those, then the average wait time is 49 days. (Uncanny! That's exactly what our in-thread estimate has been!)

Q: Is NVC picking on either CSC or VSC cases predominantly?

A: The 36 cases are composed of 23 CSC, (64%) and 13 VCS (36%). Again, if we discard the early 2007 cases the breakdown is 61% CSC, 39% VSC. Somewhat more CSC, but both USCIS centers are well represented in the "stuck in AP at NVC" status. I don't think they're picking on one center more than the other.

Q. What beneficiary countries are getting stuck? Are certain countries likelier to get pulled into NVC AP?

A. It doesn't seem that way. There are 1st and 3rd world countries represented, as follows (parenthesis indicates number of cases):

Albania (1) Australian (1) Brazil (1) Canada (1) China (4 - these include the probable duplicates)

Colombia (1) Croatia (1) Egypt (1) India (1) Mexico (1) New Zealand (1) Nigeria (1) Philippines (9)

Romania (1) Russia (2) Rwanda (1) South Korea (1) Sweden (1) Ukraine (3) United Kingdom (2) Vietnam (1)

Philippines has a high number but I think it has the majority of K-1 cases to begin with (at least on VJ)

Q. After finally getting released from NVC hell, how long will it take till my petition gets to the Consulate?

A. The answer seems to be: quickly: If we discount the early 2007 records, the average is 7 days from "NVC Left" until "Consulate Received". The early 2007 records had counts > 100 days, which is partly why I suspect that data.

Q. Can I see the names of those who got stuck? And see where I am on the "all time list"?

A. Yup. See the first attachment. You'll see some common names there like badgerella (51 days) and me&him (that's cdneh) at 52 days. They're not the "winners" however: take pity on poor Andy&Kaitlyn (98 days) or Misty&Eric (83 days). However all those date back to 2007 or early 2008. Everyone who's gotten stuck in Nov/Dec 2008 has been released in 52 days or less (Yes, cdneh does hold that record).

Q.What about K3?

A.. Sorry, I haven't analyzed K3's any further. I think the data will be similar, but (a) there's less of it to do a thorough survey (b ) I'm a K-1 so that's where my interest lies, sorry. If you want to see K3 data, let me know and I can take a crack at it.

Q. I don't like text and numbers. Don't you have any pictures to show me?

A. Yup. I've attached a chart showing 2007-2008 data. It clearly shows the dramatic spike up in Nov/Dec for the number of files held up in AP at NVC. Especially if you toss out the Jan-07 bar, it's pretty obvious SOMETHING NEW is going on at NVC.

Q. What does this all mean anyway?

A. 42

Q. Why did you do this?

A. I was pretty bored I guess. Actually, I am planning to write a letter to NVC with this data and demand some explanation for what is going on there. And if I don't get satisfaction from them, I think I'll submit this to my 2 US Senators and my Congressman. Somebody in authority needs to be aware of this.

Q. Will it do any good?

A. Of course not.

[Talyss 0]

awesome further analaysis

Very nice that you were able to produce time comparisons for people *big round of applause*

I really hope you have good news soon

[cdneh 3,112]

I think it's a remarkable piece of work! Well done!

[Badgerella 5]

Wow, well done, again!

[queuedup 1]

That's definitely a nice piece of work!!

[toddandhien 2]

OK, here is some more analysis.

I did several more things.

First, I went back to all the data for the past two years - 2007 and 2008.

I also played around with the threshold, looking at NVC times of > 10 days, 15 days, 20 days, 40 days.

In all cases, the trend of a spike up in Nov, Dec 2008 is apparent. However it is most dramatic when looking at > 40 days, so that's what I'm presenting here.

It turns out there are 36 K-1 applications over the past 2 years 2007 and 2008 which took > 40 days at NVC

These 36 are the "stuck in NVC AP" cases.

22 of these 36 (61%) occurred in the 2 months of Nov/Dec 2008.

Also, there is duplication and I believe suspect data in the earliest records from 2007. Four of those records are reported from China and I believe are duplicates. Tossing those means 22/33 = 67% are in the last 2 months. In any event it's clear SOMETHING is going on at NVC recently.

Q: What is the average wait time in NVC for those who get stuck?

A: The average wait time for all 36 is 62 days. However the suspect records from early 2007 had wait times > 100 days. If we don't count those, then the average wait time is 49 days. (Uncanny! That's exactly what our in-thread estimate has been!)

Q: Is NVC picking on either CSC or VSC cases predominantly?

A: The 36 cases are composed of 23 CSC, (64%) and 13 VCS (36%). Again, if we discard the early 2007 cases the breakdown is 61% CSC, 39% VSC. Somewhat more CSC, but both USCIS centers are well represented in the "stuck in AP at NVC" status. I don't think they're picking on one center more than the other.

Q. What beneficiary countries are getting stuck? Are certain countries likelier to get pulled into NVC AP?

A. It doesn't seem that way. There are 1st and 3rd world countries represented, as follows (parenthesis indicates number of cases):

Albania (1) Australian (1) Brazil (1) Canada (1) China (4 - these include the probable duplicates)

Colombia (1) Croatia (1) Egypt (1) India (1) Mexico (1) New Zealand (1) Nigeria (1) Philippines (9)

Romania (1) Russia (2) Rwanda (1) South Korea (1) Sweden (1) Ukraine (3) United Kingdom (2) Vietnam (1)

Philippines has a high number but I think it has the majority of K-1 cases to begin with (at least on VJ)

Q. After finally getting released from NVC hell, how long will it take till my petition gets to the Consulate?

A. The answer seems to be: quickly: If we discount the early 2007 records, the average is 7 days from "NVC Left" until "Consulate Received". The early 2007 records had counts > 100 days, which is partly why I suspect that data.

Q. Can I see the names of those who got stuck? And see where I am on the "all time list"?

A. Yup. See the first attachment. You'll see some common names there like badgerella (51 days) and me&him (that's cdneh) at 52 days. They're not the "winners" however: take pity on poor Andy&Kaitlyn (98 days) or Misty&Eric (83 days). However all those date back to 2007 or early 2008. Everyone who's gotten stuck in Nov/Dec 2008 has been released in 52 days or less (Yes, cdneh does hold that record).

Q.What about K3?

A.. Sorry, I haven't analyzed K3's any further. I think the data will be similar, but (a) there's less of it to do a thorough survey (b ) I'm a K-1 so that's where my interest lies, sorry. If you want to see K3 data, let me know and I can take a crack at it.

Q. I don't like text and numbers. Don't you have any pictures to show me?

A. Yup. I've attached a chart showing 2007-2008 data. It clearly shows the dramatic spike up in Nov/Dec for the number of files held up in AP at NVC. Especially if you toss out the Jan-07 bar, it's pretty obvious SOMETHING NEW is going on at NVC.

Q. What does this all mean anyway?

A. 42

Q. Why did you do this?

A. I was pretty bored I guess. Actually, I am planning to write a letter to NVC with this data and demand some explanation for what is going on there. And if I don't get satisfaction from them, I think I'll submit this to my 2 US Senators and my Congressman. Somebody in authority needs to be aware of this.

Q. Will it do any good?

A. Of course not.

This is real good work. Well done.

Regarding the bolded Q/A above, 61/39 would appear significant. However, if you could look at the data and assess the overall CSC/VSC ratio and compare that to the 61/39, you could get a real assessment of whether they are picking on one of the service centers. I suspect just by looking at the postings here, there are more folks going through CSC than VSC, so probably OK.

[Y's_habibitk 2]

All right gang, well I finally got around to compiling some stats on our NVC wait. I'd been meaning to do this for a few weeks.

To show you just how PATHETIC is the life of a guy waiting for his honey's visa so they can finally be reunited ..... yes, I actually am spending my Valentines night doing this!

How's that for a guy with literally no life???

Some technical details, for those so inclined...

First of all - there is A LOT of data in the Visajourney database. As we all knew by looking at Igor's page, it is possible to get great stuff from this database. Igor has obviously mastered the ability to run a cron-based job to automatically pull the dataset hourly. He seems to run SQL queries directly against the VJ database, Perhaps he's figured out the VJ http syntax to format his queries via the webserver but it doesn't look to me as though that's what he's doing.

Anyway, I don't have direct SQL access to the db and I'm not smart enough to decipher the http parameter strings. So I just hacked it manually.

Instead I just used the built in query builder on the VJ website itself. That kind of sucks because it's pretty weak, I can't run the queries I'd really want to. It lets you hide columns, but the row filtering was not helpful. I just grabbed all rows in the database for which "NVC Received" was non-null. That was 7113 rows out of a total 16,310 in the database. After playing around to narrow my data range as best I could, I just grabbed the resulting html pages manually and cut/pasted the raw data into Excel.

I then spent hours tonight working with Excel DCOUNTA and DAVERAGE functions to get the stats I wanted. Excel is garbage, imho. I should have done it in all in Perl, it would have taken me far less time. I don't know what possessed me to do it in Excel. Oh well.

Unfortunately VJ does not have a field on our timeline data for "AP at NVC". So it's not possible to know explicitly which applications got stuck in NVC AP.

What our timelines do have is the Date Entered NVC and Date Left NVC. Even then, unfortunately, not all VJ members recorded both of these fields. In many cases they record only one or the other or neither, so I can't compute the time spent at NVC.

So what I did was compute the time spent at NVC for those who recorded both values.

The earliest record with 'NVC received' date in the database is 8/23/1995 That's got to be a typo. The 'NVC left' date for that record was in 2008. What, they spent 13 YEARS in NVC??? Um, I doubt it. So ok, that prompted me to toss all records for which NVC Left-NVC entered was too big (I picked 1000 days, that threw away all the junk like this one).

I then compared the time spent at NVC to a threshold value which I felt was large enough to indicate that any application which spent that long at NVC probably got stuck in AP. I picked the threshold at 15 days.

That's important to remember!!! If you pick a different threshold (10 days, 20 days), you'll get different results! Caveat emptor.

The data generally covers the period 2003-2009, which seems to be the period VJ has been operational. There are some data from < 2003, but very few, and I suspect they are typos. As you'll see I focused most of my attention on the past year, 2008.

Anyway, based upon a 15 day threshold I then computed break-down stats for K1 and K3 filings, and I did that month-by-month for 2008 to see if things trended worse at the end of last year. Guess what - no surprise to all of us stuck at NVC, they ARE trending worse!!!

OK, here are the results:

Total number of rows in VJ database: 16,310

Number of rows that have a 'Start NVC' date: 7,113 (=== this is the total number of raw data rows imported into Excel spreadsheet)

Number of rows that are K1s : 5,002

Number of rows that are K3s: 836

Number of rows that are IR1/CR1: 1,275

Well, that's not really so interesting. We really care about rows that have BOTH start and end NVC dates:

Number of records that have start/end NVC dates 4,696

Number of records that don't have start/end NVC dates 2,417

Number of K1s that have start/end NVC dates 3,520

Number of K3s that have start/end NVC dates 603

Number of IR1/CR1s that have start/end NVC dates 573

OK, now here things start to get interesting. These are the average times (in days) for each visa category spent at NVC over the entire database history (2003-2009):

Average time in NVC for all records 20.349

Average time in NVC for K1s 6.918

Average time in NVC for K3s 10.121

Average time in NVC for IR1/CR1s 113.621

As we always knew, K1's generally get in and out quickly - an average of 6.9 days. K3s are a bit longer (10 days), and IR1/CR1 stay MUCH longer because that's where most of their processing occurs.

Right. Next.

Here is the K1 data, broken down month by month since Jan 08:

               Num K1/month    Avg NVC time    Num K1 > threshold    Avg NVC time > threshold    % K1 in month > threshold
Jan-08    111                    6.414            7                            35.4286                            6%
Feb-08    126                    5.651            7                            26.2857                           6%
Mar-08    125                    4.744            5                            21.2000                           4%
Apr-08    56                    4.714            4                            19.7500                           7%
May-08    78                    4.795            2                            28.5000                           3%
Jun-08    46                    4.761            1                            32.0000                           2%
Jul-08    95                    4.137            4                            19.0000                           4%
Aug-08    58                    6.448            4                            31.0000                           7%
Sep-08    71                    5.577            6                            24.3333                           8%
Oct-08    69                    4.710            3                            26.6667                           4%
Nov-08    64                    10.125            11                            43.2727                           17%
Dec-08    73                    12.452            15                            44.0000                           21%
Jan-09    116                    3.845            0                            #DIV/0!                           0%
Feb-09    13                    2.231            0                            #DIV/0!                           0%

Here's how to interpret this. Look at the first row for Jan 08, for example. In that month there were 111 K1 visas that entered NVC and also exited at some point. Those 111 applications spent an average of 6.4 days at NVC. Seven of those visas spent 15 days or longer at NVC, and they each spent an average of 35.4 days at NVC - good chance that they were 'AP' cases. 7/111 is 6%, so that's approximately the percentage of all Jan08 K1s which got stuck in AP.

Ignore the Jan/Feb 09, since those contain records which have entered NVC but not yet exited > 15 days ... That's US FOLKS!!!! STUCK NOW AT NVC!!! Hence those rows have 'zeroes' for Num K1 > threshold, and #Div/0 for the avg time > threshold.

OK, so now if you look down through the months we see something disturbing. In Nov and Dec 2008, BOTH of the last two columns jump up.

Suddenly in Nov/Dec the average time that the 'long' cases (those > 15 days which we suspect are AP) are spending an average of 43 or 44 days at NVC which is considerably higher than previous months. And it gets close to our current anecdotal evidence on this VJ thread that the wait is about 50 days.

Also, we see in Nov/Dec that 17% and 21% , respectively, of all K1 cases are waiting more than 15 days and so are suspect of being AP.

That tells us two things:::::

There seem to be MORE AP cases happening. And when they happen, they are stuck in AP longer.

Sigh.

Well, at least now we know it's not our imagination. It's really happening.

I did a similar analysis for the K3's but since there's much less data, the statistics are less relevant (law of large numbers etc.)

I've attached 2 charts showing this end of year trend in the K1 data.

I tried to attach my Excel file with all the data - all raw data, calculations, and the charts of the K1 2008 data. However VJ does not seem to allow me to update .xls files.

If you would like to see this file, just PM me and I will email it to you.

Happy Valentines all.

Looking forward to getting out of NVC HELL

uscandual...you f'ing ROCK.

thank you for putting all of that effort in and sharing it with us

Lisa

[scandal 951]

Q: Is NVC picking on either CSC or VSC cases predominantly?

A: The 36 cases are composed of 23 CSC, (64%) and 13 VCS (36%). Again, if we discard the early 2007 cases the breakdown is 61% CSC, 39% VSC. Somewhat more CSC, but both USCIS centers are well represented in the "stuck in AP at NVC" status. I don't think they're picking on one center more than the other.

This is real good work. Well done.

Regarding the bolded Q/A above, 61/39 would appear significant. However, if you could look at the data and assess the overall CSC/VSC ratio and compare that to the 61/39, you could get a real assessment of whether they are picking on one of the service centers. I suspect just by looking at the postings here, there are more folks going through CSC than VSC, so probably OK.

You're quite right! Good eye.

I implicitly assumed that there are more CSC than VSC in the general population, and hence I glossed over the 61/39 ratio.

So, here is the actual breakdown of service centers in all of the 2705 records I'm analyzing for 2007/2008:

1479 California

990 Vermont

112 Nebraska

124 Texas

2705 TOTAL

As an aside .... can anyone explain what the Nebraska and Texas centers are doing here? I thought all K1 files are handled at either California or Vermont?

However if we ignore those, than the relative proportions of CSC and VSC to the general population of records is as follows:

CSC: 1479/(1479+990) = 60%

VSC: 990/(1479+990) = 40%

OK. CSC/VSC is about a 60/40 split in the general population, so NVC AP stats of 61/39 don't reflect any bias.

To all who have offered thanks.... aw shucks tweren't nothin'... seriously it was either do this, or work on my tax return

[toddandhien 2]

Q: Is NVC picking on either CSC or VSC cases predominantly?

A: The 36 cases are composed of 23 CSC, (64%) and 13 VCS (36%). Again, if we discard the early 2007 cases the breakdown is 61% CSC, 39% VSC. Somewhat more CSC, but both USCIS centers are well represented in the "stuck in AP at NVC" status. I don't think they're picking on one center more than the other.

This is real good work. Well done.

Regarding the bolded Q/A above, 61/39 would appear significant. However, if you could look at the data and assess the overall CSC/VSC ratio and compare that to the 61/39, you could get a real assessment of whether they are picking on one of the service centers. I suspect just by looking at the postings here, there are more folks going through CSC than VSC, so probably OK.

You're quite right! Good eye.

I implicitly assumed that there are more CSC than VSC in the general population, and hence I glossed over the 61/39 ratio.

So, here is the actual breakdown of service centers in all of the 2705 records I'm analyzing for 2007/2008:

1479 California

990 Vermont

112 Nebraska

124 Texas

2705 TOTAL

As an aside .... can anyone explain what the Nebraska and Texas centers are doing here? I thought all K1 files are handled at either California or Vermont?

However if we ignore those, than the relative proportions of CSC and VSC to the general population of records is as follows:

CSC: 1479/(1479+990) = 60%

VSC: 990/(1479+990) = 40%

OK. CSC/VSC is about a 60/40 split in the general population, so NVC AP stats of 61/39 don't reflect any bias.

To all who have offered thanks.... aw shucks tweren't nothin'... seriously it was either do this, or work on my tax return

Wow, that worked out even closer than I had hoped. That tells me that the AP selections are truly random, which would work out to statistically mimic the population over time. Thanks for doing that. Next valentine's day you will have something else to occupy your time...

[scandal 951]

Wow, that worked out even closer than I had hoped. That tells me that the AP selections are truly random, which would work out to statistically mimic the population over time. Thanks for doing that. Next valentine's day you will have something else to occupy your time...

Well, you inspired me to verify the assertion that the by-country selections are also truly random and show no overall bias.

I think that's true, but have a look and tell me what you think.

See the attached table for reference (I can't for the life of me get VJ's html editor to properly format my tables, so I've taken to freezing them as a jpg and attaching them).

As you can see, Philippines has 547 K1 applications, which is roughly 20% of the total. Hence we'd expect 20% of the 36 (which is 7.3) NVC-AP cases we are studying to be Philippines. In fact, 9 cases were Philippines. So, that's reasonably close.

Ukraine has an expectation of 1 NVC AP case, in fact it had 3. Bum luck, Ukraine.

Canada should have had 2 NVC AP cases by now, it's had only one. Lucky Canucks.

Vietnam is batting par - expect 0.8, got one (although with one more on the way, Todd and Hien).

Thailand expects 1.1, has zip so far, but I'm waiting my turn to make prediction match reality.

In short, it all seems about right.

NVC is just randomly picking cases for AP, at an apparent rate of 14-18% of overall K1 cases for the months of Nov and Dec 2008. It looks like this trend is continuing in Jan 2009 and forward. And I think it applies equally to K1s and K3s (unsubstantiated, but looks that way).

I'm guessing they got an internal directive to pull about 15-20% of cases at random.

i certainly hope than by Valentines Day 2010, ALL OF US have much more interesting things to do with our loved ones than hang out on VJ

[Talyss 0]

I think Nebraska and Texas are in there as it was sometime in 2007 they changed to it being only 2 centres that processed K1s before that all 4 did.