I downloaded the severance agreement. Then, I was able to grab the categories and the ages of each employee from the bottom of the document. I verified that my age was on the list along with my specific category. Then I copied the data into a csv file and uploaded it into an AI (copilot). The AI performed the calculations quickly yielding that 77% of employees older than 40 were let go of the less than 2k employees in total on this rif (reduction in force). In case anyone wishes to verify my numbers, that is why I just posted this. This took me less than 5 mins to do just so you know the type of employee that I was and that I was let go. Also, if anyone else gets a severance docusign agreement and wishes to pursue a legal age discrimination action, don't sign that agreement. If you do, you may lose your right to sue.
9 replies (most recent on top)
@bs Are you making an assumption based on observation or a fact-based one? In other words, what is your source of data that you are making that claim based on?
@bs The only way to accept what you are saying is to know the overall population of the data. That is, one would need to know the entire number of over 40 versus under 40 employee breakdowns within the entire corporation. Or perhaps just within the division. Given the company hasn't split yet, the entire corporation's employee database would need to be used in my opinion. Then, they could calculate if what you said is correct.
Statistics often require making sample size projects to the population. Given the 77% let go within my severance agreement (HW provided that data to me at the bottom of the document so that is a given), and given that the sample size of those let go were 2% of the entire 102K population of employees, this is
Here is what the Copilot AI stated when fed this situation:
Sample Size Adequacy
Population size: 102,000 (Honeywell employees)
Sample size: 2,000 (those laid off)
Proportion over 40: 77% of the sample
To determine if the sample is “large enough,” statisticians often use the Central Limit Theorem, which says that for proportions, a sample size of at least 30 is often sufficient for normal approximation — but for more robust inference, samples over 1,000 are considered very strong.
So yes, 2,000 is statistically large enough to estimate proportions with high confidence.
Bottom Line
Yes, 2,000 is a statistically valid sample size.
The test you're referring to is a z-test for proportions.
If 77% of those laid off were over 40, and that’s significantly higher than the company’s age distribution, it could indicate disparate impact — a key concept in employment discrimination law.
Would you like help running the actual z-test with assumed age distribution data?
Here’s the result of the z-test for proportions based on your scenario:
📊 Z-Test Summary
Sample size (n): 2,000 laid-off employees
Observed proportion (p̂): 77% were over 40
Expected proportion (p₀): 50% (assumed company-wide age distribution)
🧮 Results
Z-score: 24.15
P-value: < 0.0001
✅ Interpretation
This result is highly statistically significant. A z-score of 24.15 means the observed proportion (77%) is 24 standard deviations away from the expected (50%) — an enormous deviation. The p-value being effectively zero confirms that this difference is not due to random chance.
In plain terms:
The age distribution of those laid off is not representative of the general workforce. It strongly suggests a disparate impact on older employees.
@gv I believe you should be able to file for unemployment insurance. I know that I just did but I'm waiting on the state to approve it. You should check with your state and also just ask an AI to give you its view.
I'm card assembler and very worry. I need job. Do I get government payment if I get lay off?
All severances veteran’s contact information was likely sold to Ukraine, Russian, Israel, Pakistan, and India just to help in job placement. So helpful.
What division and location?
Who got hit in this round?
Keep in mind, the workforce in HW is skewed to older workers. We have trouble recruiting younger talent. So it may not be age discrimination so much as we have an older workforce period. I work with only one person under 40.
Did they list how many veterans?