3 ms·

It's also a great example of using statistics to create a conclusion that fits a narrative, based upon statistically insignificant behavior. This article is en

by IncRnd 8y ago

It's also a great example of using statistics to create a conclusion that fits a narrative, based upon statistically insignificant behavior. This article is entitled, "Over the Past 5 Years, Inflation for Millennials was 14% Higher than for Baby Boomers". Yet, we find out in the last paragraph, indeed in the final sentence, that there is virtually no difference between groups over a longer, more significant period of 18 years. There are three kinds of lies: lies, damned lies, and statistics.

tjpaudio 8y ago
Their use of the word significantly in this context does not refer to statistical significance. The percentages you are referring too are just basic metrics based on the last years data, not something that was t-tested. The longer time period changes the starting index, it doesn't really make this type of analysis more or or less statistically significant. As the author alluded, this could mean that economic mechanisms close this gap over time when it appears, or it could be a recent effect. Your criticism isn't really valid.
- IncRnd 8y ago
  You are making a statement about statistics that isn't correct. Statistics is collecting and analyzing numerical data, which is what this article does. That is a truth, regardless of it being a different sort of analysis than what you believe to be the only model of statistics. That is also orthogonal to what I wrote, which you seemed to intentionally misunderstand in order for you to argue.
CompelTechnic 8y ago
I think the way that percentages were expressed in the headline was intentionally deceptive- 14% seems like a lot of inflation in 5 years! That makes for good clickbait. But the percent that is being talked about is the comparative difference between groups, per the line: Over the past five years, millennials have seen their goods become 8.7% more expensive, Generation X 8%, and baby boomers 7.6%. This means that the price increase for millennials was 14% higher than for baby boomers.
- IncRnd 8y ago
  I think the way that percentages were expressed in the headline was intentionally deceptive- 14% seems like a lot of inflation in 5 years! That makes for good clickbait. Yep. That's my point. The article's premises do not support the conclusion. And, in actuality the end of the article contradicts its own conclusion. But the percent that is being talked about is the comparative difference between groups, per the line: Over the past five years, millennials have seen their goods become 8.7% more expensive, Generation X 8%, and baby boomers 7.6%. This means that the price increase for millennials was 14% higher than for baby boomers. There is a subtle error in this analysis. If companies A and B both grow with A growing 300% and B growing 1% per year, it seems that A might be a better company. That is often why companies that don't make money are valued similarly to companies that do make money. Putting the perception aside, A may have only increased by $4 but B could have increased by $4,000,000,000. Therefore it is not correct to say that the increase of A was 29,999% higher than for B.
- seandougall 8y ago
  Comparing percentages always has the potential to invoke that sort of confusion, but FWIW I'm pretty sure the headline is written correctly. If it were subtracting percentage values from one another rather than dividing, it would be "percentage points" rather than "%". That said, the article itself is just silly, a true puff piece. I think one could make a more substantive (and more interesting) argument around the idea of the "basket of goods" measurement as a basis for measuring inflation in the first place, as goods make up a smaller and smaller percentage of our total expenses.

by IncRnd 8y ago

tjpaudio 8y ago

Their use of the word significantly in this context does not refer to statistical significance. The percentages you are referring too are just basic metrics based on the last years data, not something that was t-tested. The longer time period changes the starting index, it doesn't really make this type of analysis more or or less statistically significant. As the author alluded, this could mean that economic mechanisms close this gap over time when it appears, or it could be a recent effect. Your criticism isn't really valid.

IncRnd 8y ago
You are making a statement about statistics that isn't correct. Statistics is collecting and analyzing numerical data, which is what this article does. That is a truth, regardless of it being a different sort of analysis than what you believe to be the only model of statistics. That is also orthogonal to what I wrote, which you seemed to intentionally misunderstand in order for you to argue.

CompelTechnic 8y ago

I think the way that percentages were expressed in the headline was intentionally deceptive- 14% seems like a lot of inflation in 5 years! That makes for good clickbait. But the percent that is being talked about is the comparative difference between groups, per the line: Over the past five years, millennials have seen their goods become 8.7% more expensive, Generation X 8%, and baby boomers 7.6%. This means that the price increase for millennials was 14% higher than for baby boomers.

IncRnd 8y ago
I think the way that percentages were expressed in the headline was intentionally deceptive- 14% seems like a lot of inflation in 5 years! That makes for good clickbait. Yep. That's my point. The article's premises do not support the conclusion. And, in actuality the end of the article contradicts its own conclusion. But the percent that is being talked about is the comparative difference between groups, per the line: Over the past five years, millennials have seen their goods become 8.7% more expensive, Generation X 8%, and baby boomers 7.6%. This means that the price increase for millennials was 14% higher than for baby boomers. There is a subtle error in this analysis. If companies A and B both grow with A growing 300% and B growing 1% per year, it seems that A might be a better company. That is often why companies that don't make money are valued similarly to companies that do make money. Putting the perception aside, A may have only increased by $4 but B could have increased by $4,000,000,000. Therefore it is not correct to say that the increase of A was 29,999% higher than for B.
seandougall 8y ago
Comparing percentages always has the potential to invoke that sort of confusion, but FWIW I'm pretty sure the headline is written correctly. If it were subtracting percentage values from one another rather than dividing, it would be "percentage points" rather than "%". That said, the article itself is just silly, a true puff piece. I think one could make a more substantive (and more interesting) argument around the idea of the "basket of goods" measurement as a basis for measuring inflation in the first place, as goods make up a smaller and smaller percentage of our total expenses.