Food Inflation Experiment +

[Virgil Showlion]

Welcome to the Food Inflation Experiment (FIE) +, originally created by midwesterner and now being administered by Virgil Syonid.

The FIE tracks changes to food prices across North America using an application in the Market Talk Application Bar.



Participants:

• Please see Reply #1 in this thread for instructions on how to view your data.
• Please see Reply #2 in this thread for instructions on how to enter your data.
• Please see Reply #3 for a summary of the experiment rules and FAQs.

Everyone:

• Please see Reply #1 in this thread for instructions on how to view the experiment data.
• Please contact Virgil or post in this thread if you have any (i) difficulties viewing the data, (ii) questions about the experiment, (iii) requests for raw copies of the data, or (iv) general questions/concerns.

---

This thread will remain an announcement until 12:00 AM April 8, 2011.

Participants will be notified by PM that the experiment is up and running.

Regards,

Virgil Syonid

[Virgil Showlion]

Viewing Data

• All data is viewable from the Food Inflation Experiment application in the Market Talk Application Bar at the top of the Market Talk board page (Fig. 1).
  
  
  Fig 1. FIE on MT App Bar
  
  If the application tab is not open, it can be expanded by clicking on the '+' symbol at the bottom of the tab.
  
  
• Participants will see Enter Data and FIE Thread buttons at the top right of the application (Fig. 2). Non-participants will see only the FIE Thread button.
  
  
  Fig 2. FIE Buttons
  
  The Enter Data button will open up the data entry interface.
  The FIE Thread button, which is always available, will redirect the user's browser to this thread.
  
  
• The data selection dropdown box at the top left of the application (Fig. 3) allows users to select the data they wish to view.
  
  
  Fig 3. Data Selection Dropdown Box
  
  The viewable datasets are (in order):
  • All Regions
  • Individual US/Canadian Regions
  • Individual Participants (Alphabetical By Display Name)
  
  
  
• The 'All Regions' dataset and US/Canadian Regional datasets are aggregated over all participants in the experiment (All Regions), or in specific regions (Regional).
  
  
• Each dataset consists of prices for 10 food items (see Reply #3 for details) for 12 months in 2011 (Fig. 4).
  
  
  Fig 4. Price Data
  
  Any column (month) that has all prices entered will also have a total price (sum of all prices) automatically computed. The % Data Complete row at the bottom of the table shows what percentage of the data is available for each month. Only complete records (months with 100% data completion) are counted toward the 'All Regions' and Regional datasets.
  
  
• Each complete month (column) will also display the following data:
  • % Change (MoM) - shows the percentage increase/decrease in total price since the previous month. This datum is only available for a given month if all data is also available for the previous month.
  • % Change (Total) - shows the percentage increase/decrease in total price since January. This datum is only available for a given month if all data is also available for the month of January.
  • % Change (Ann.) - shows the annualized percentage increase/decrease in total price since January (the best estimate of yearly inflation). This datum is only available for a given month if all data is also available for the month of January.
  
  
• When viewing aggregate data (such as 'All Regions'), the % Data Complete row (see Fig. 4) shows the percentage of participants that have submitted complete records for each month.
  
  
• When viewing aggregate data, the Total price is the average of all Total prices for complete records.
  
  The % Change (MoM), % Change (Total), and % Change (Ann.) data are computed in two different ways for aggregate data, depending on whether the 'Normalize Data' button (Fig. 5) is checked.
  
  
  Fig 5. 'Normalize' Button
  
  When data normalization is on, each percentage reflects the average of all computed percentages. A participant's percentage for a month is only included in this average if the percentage is viewable. For example, if user 'A' has complete data for March but not February, he/she will not have any viewable % Change (MoM) data for February and hence will not contribute to the February average.
  
  When data normalization is off, each percentage reflects the percentage associated with average total cost. As when normalization is on, a participant's data is only used in the computation of an aggregate percentage if the percentage is viewable in his/her data.
  
  As as example, suppose we have participants:
  • A: [Jan: $100, Feb: $110]
  • B: [Jan: $1000, Feb: $1000]
  • C: [Jan: ?, Feb: $10000]
  • D: [Jan: $50, Feb: ?]
  
  In this case, only A and B are included in the aggregate percentages for February because C is missing January data and D is missing February data.
  
  The application will report 75% data complete for January; 75% complete for February.
  
  The total cost for January will be reported as ($100 + $1000 + $50)/3 = $383.33.
  
  The total cost for February will be reported as ($110 + $1000 + $10000)/3 = $3703.33.
  
  In normalization is on, the application will report:
  • % Change (MoM): (($110 - $100)/$100 + ($1000 - $1000)/$1000)/2 = 5%
  • % Change (Total) = 5%
  • % Change (Ann.) = (1 + 5%)^12 - 1 = 79.6%
  
  If normalization is off, the application will report:
  • % Change (MoM): (($110 + $1000) - ($100 + $1000))/($100 + $1000) = 0.9%
  • % Change (Total) = 0.9%
  • % Change (Ann.) = (1 + 0.9%)^12 - 1 = 11.5%
  
  
  In both cases, the calculated percentage is vastly different from the amount obtained by dividing the February average by the January average.
  
  If food prices do not vary markedly between regions and/or if many participants submit data, the two methods will report similar percentages.
  
  The Normalize Data button has no effect when viewing individual participant data.

[Virgil Showlion]

Submitting Data

• Participants can submit data to the FIE at any time by clicking on the Enter Data button (see Fig. 2 in Reply #1). This will open up Data Entry Mode (Fig. 6) in the FIE application.
  
  
  Fig. 6. Data Entry Mode
  
  The columns and rows are identical to View Mode, with the exception that (i) the rows for the total, percentage, and completion data are not shown, and (ii) the values in the table become editable.
  
  
• Administrators are given a dropdown box in the top left of the application (see Fig. 6) to select the dataset they wish to edit. Participants will see only static text indicating that they are editing their own dataset.
  
  
• Any price in the table can be edited by clicking on it (a box appears while hovering over table cells to indicate which cell is being targeted).
  
  A price can be typed in with or without a dollar sign ($) and with or without a decimal place or leading zero. To finish editing a cell, either (i) use the arrow keys to move to another cell, (ii) hit Enter, (iii) hit Escape to return the cell to its original value, (iv) click in another cell to edit it, or (v) click on something outside of the cell.
  
  The price in a cell will be formatted automatically when the cell loses focus. If the price entered exceeds $100, the price will be clipped to $99.99. If the price entered is not valid, the cell will revert to its original value.
  
  Note that the contents of a cell can be deleted if desired.
  
  
• As soon as one or more changes have been made to the price table, the Save Changes button (Fig. 7) becomes active.
  
  
  Fig. 7. 'Save Changes' Button
  
  For changes to the price table to be committed to the database, participants must click on the Save Changes button. If the button is not clicked, no changes will be stored in the database.
  
  Participants can edit and save data as many times as needed.
  
  
• The View Data button will return the application to Data View mode. If there are unsaved changes to the price table, a dialog will pop up to request if these changes should be saved to the database.
  
  Note that saving data will automatically set the active dataset in View Mode to the saved dataset. For example, saving data for 'ArchietheDragon' will automatically show Archie's information when the application is returned to View Mode.
  
  
• Participants who are unable or unwilling to enter data manually can alternatively post their data in this thread or send it via PM to Virgil.
  

[Virgil Showlion]

Rules and FAQ

• The original Food Inflation Thread is viewable here. It was locked when this alternate thread was created.
  
  
• The regions in the experiment include Canada (as a single region) and the USA subdivided as follows:
  
  
  
  
• Experiment Rules as given in the original FIE:
  • No substitutions allowed for food items unless the substitution is noted in this thread.
  • Product quantities must be within +/-10% of the indicated quantities.
  • No discount retailers can be used. All stores must be 'average' grocery stores that do not sell in bulk or require a membership. Sam’s Club and Costco are not allowed.
  • The same grocery store and same food items must be used for an entire year.
  • Participants are encouraged to submit data every month (see below for details).
  • No coupon prices or in-store promotional prices are allowed unless explicitly noted in this thread.
  • Prices should reflect the sticker price of a product, excluding bulk discounts and any applicable sales tax.
  • Participants are kindly asked to be diligent, honest, and careful in their reporting.
  
  
• Participants may receive occasional PMs (notices, reminders, etc.) in regards to the FIE.
  
  
• The original FIE had six two-month periods rather than twelve monthly periods. Participants that are unwilling or unable to move to monthly reporting should place two-month data in even month columns (e.g. February, April, June, August, October, December).
  
  All participants are encouraged to submit price data for all months.
  
  
• January data is still missing for several participants. Please submit this data ASAP.
  
  
• Any questions, concerns, problems, etc. can always be send via PM to Virgil or posted in this thread.
  

A warm "thank you!" to all participants!

[Virgil Showlion]

Daphne, Mid and I are still figuring out if there's an elegant way of bringing in new participants mid-year. Normally an experiment like this would require January data as a baseline, but there are still a few options available to us.

I shall keep interested parties apprised.

Thanks for volunteering.

[decoy409]

NICE JOB VIRGIL!!!

Of course with the variations in price as time goes on we also see weaking in prices in certain respects in order to hold up sales of product. Would this suggest that some companies are not as strong as thought when push comes the shove in a deflation/inflation state of things?

Of course there will be a negative reaction by the one that wants to be captain as this may hamper 'some' of those great calls.

[usaone]

Looks like food prices are dropping nicely.

They increased because of weather not the Fed.

[decoy409]

Thanks for the GREAT LAUGH above this morning!

Quote: Looks like food prices are dropping nicely.

[rovo]

Apr 7, 2011 9:25:18 GMT -5 decoy409 said:

Thanks for the GREAT LAUGH above this morning!

Quote: Looks like food prices are dropping nicely.

Decoy, Enough with snarky comments or I'll toss your ass right off the boards. If you are disputing the data then state where the error is, else shut the hell up.

[lifewasgood]

You got to be kidding Rovo

[rovo]

Apr 7, 2011 9:34:15 GMT -5 @lifewasgood said:

You got to be kidding Rovo

\

Nope. Not kidding. Think I am? Try me.

[lifewasgood]

Try what?

[jarhead1976]

Rovo Sir, Its been obvious before, that decoy and mid upset you. A little patience goes further than any threat. Without debate there is only one point of view. I see decoys remark as being no more "snarkey" than Frank the Impalers. Lets leave the bias and hostility out. I know you Mods and administrators work hard to keep the level of hostility under control . Thank you for all your hard work . I think this Thread and the results will be worth the effort. They should be respected by everyone . Respectfully. jarhead1976

[decoy409]

Region U.S. Southwest - 6 out of 10 items have INCREASED

Region U.S. West - 10 out of 10 items have DECREASED

Now have a look at, www.latoyaegwuekwe.com/geographyofarecession.html

Then compare the following U.S. Regions noted above as these are the only tales of the tape so far to show the respectable balance they play as to the Unemployed and the rise or depletion of food costs affecting those.

Next all one has to do is to see where those are at that entered the information as to the prices,and where they are located.

And after have a look at the highest states of unemployed. And from that we can form a TREND as to things are getting better or worse for these people in their respectable states.

Now is there a problem with that?

[lifewasgood]

Yes, this is not a controlled experiment, with no or little verification.

[frankq]

Apr 7, 2011 9:25:18 GMT -5 decoy409 said:

Thanks for the GREAT LAUGH above this morning!

Quote: Looks like food prices are dropping nicely.

The Great Laugh is on you. The data has been supplied by POSTERS. Unless you feel that the participants of this "experiment" are members of the NWO, the data speaks for itself. Your claims of unimagimable inflation, food shortages and massive civil uprising seems to be...uh...not happenning at this time.

[frankq]

Apr 7, 2011 9:53:24 GMT -5 @lifewasgood said:

Yes, this is not a controlled experiment, with no or little verification.

You only need to walk into any supermarket in your area and compare if you're looking for verification.

[lifewasgood]

Not worth responding

[frankq]

Then don't.

[verrip1]

Virgil:

Good job!

It's clear that you spent a lot of time on this. For that alone you deserve credit. My opinion of the project hasn't changed, but, regardless, you've really done well with this. I hope that Mid is also happy with what you've done.

Again, congrats.

[Virgil Showlion]

Thanks for the feedback thus far.

After reading the "prices dropping" comment, I remind everyone that Reply #1 explains what is actually being displayed.

[txmover]

FTI started it with "Gee Virgil you must have spent a lot of your time in preparing all this code and these charts to promote a banned posters thread.

And to give it such a prominate spot on the board,"I" can see we all have our work cut out for us in an effort to emulate such behavior to be treated so special."

and NO COMMENT FROM ROVO.

Rovo however chimed in with "Decoy, Enough with snarky comments or I'll toss your ass right off the boards. If you are disputing the data then state where the error is, else shut the hell up."

FrankQ said "The Great Laugh is on you. The data has been supplied by POSTERS. Unless you feel that the participants of this "experiment" are members of the NWO, the data speaks for itself. Your claims of unimagimable inflation, food shortages and massive civil uprising seems to be...uh...not happenning at this time."

and NO COMMENT FROM ROVO.


I've been reading for over 2 years now and have never participated in the conversations however, this one is too blatant to ignore.

Rovo, WTF, I think you've just shown your ass on this one.

[lifewasgood]

Virgil, looking at the list of data input participants it appears like Jarhead 1976 is the only one filling in data on a monthly bases. With the lack of data, no conclusions can be made, in fact the experiment is flawed unless controls are in place and sufficient data is gathered.

For example products such as ground beef and eggs have widely varying prices due to quality and size. I would suggest settling with very specific products from specific suppliers to run the inflation experiment.

If I am miss reading how this works then I retract until I know better.

[Virgil Showlion]

Mover: I'll let rovo respond to your comments.

Lifewasgood:: We've had some lengthy discussions on controls in the old FIE thread. So long as participants are diligent in sticking to the same food items for a year, the inflation read should be accurate. Note that when viewing normalized (Normalize Data button is checked), the data is normalized in such a way that baseline differences in cost are factored out.

As for the data being sparse (and I agree it is), the application has only been up for a few hours now. ;D

The January/April comparison will be the first true read. This is because the January data is actually the Jan/Feb data, and the April data is actually the Mar/Apr data. Most people haven't entered their April data (myself included) because they either submitted it to the thread, they were confused about the future of the experiment (it wasn't really 'running' per se during March), or they have yet to collect it (it is only the 7th, after all).

As more people enter their data, the values will become more and more accurate.

The % Complete column gives an indication of how complete the data is for any given month.

[frankq]

"FrankQ said "The Great Laugh is on you. The data has been supplied by POSTERS. Unless you feel that the participants of this "experiment" are members of the NWO, the data speaks for itself. Your claims of unimaginable inflation, food shortages and massive civil uprising seems to be...uh...not happening at this time."

and NO COMMENT FROM ROVO.


I've been reading for over 2 years now and have never participated in the conversations however, this one is too blatant to ignore.

Rovo, WTF, I think you've just shown your ass on this one. "

And just what comment would you like tex? The problem with you guys is that you won't accept ANY data that you can't manipulate for yourselves, while accusing every business and government entity of doing just that. The data provided in the thread comes from anonymous, impartial and everyday sources. Decoy called the data into question because he believes that food inflation is more severe and has predicted massive food shortages in his threads. The prices posted over several months don't seem to support that conclusion, though food inflation undoubtedly exists. Sorry if that offends you......

[workpublic]

nice programming Virg.

[lifewasgood]

Thanks Virgil,

I know you worked hard on this. I will hold further comment until the data completed percentages come up and confidence is high.

In the meantime, I can only work with Wife data, which is always accurate as she is the only grocery getter around here. Confidence is always high on that data set or else

[decoy409]

frank q., nope, not upset at all. And I am certainly sticking to my claim. Now let's highlight our posts today here in relevance to your comment and mine. In turn you can calculate the outcome shortly down the lane this year.
And yes,I do make these calls as shown in my thread in advance.

Thanks for your suport frank q. and have a great day!

[mtntigger]

Wow! I really appreciate all the confidence that you have in the people who VOLUNTEERED contintent-wide to collect prices. It is giving me lots of warm and fuzzies and increases my desire to collect more data.
I'd like to mention that the frequency of this experiment was changed since the beginning. Originally it was every 2 months; starting in mid-March, it changed to monthly. So, it's a little difficult to travel back in time to February to collect data when it's already March. If you want, though, I'll be happy to create some February numbers so that you have confidence.

[txmover]

Maybe I should have remained silent. Frank Q, at the end of the day I have learned a lot from everyone on this board. Yes, Decoy goes a little overboard sometimes however, I think we can all agree it's not all Roses all the time. Even FTI has voiced concern and caution recently. The problem I see is that so much time is spent splitting hairs and instigating each other for minor differences, we are all missing the big picutre. The big picture is finding a solution. Dem's, GOP, Tea Party, Rat Pack, D&G's....so much time spent pissing in each others boots that we fail to realize that in the end we're all standing in the same stuff. In regard to your reply to me, I am not one of "you people" that "always" tries to manipulate. That's a big assumption drawn on a single post. It's all good in the end. Peace!!