# Thread: The Danger of Statistics

1. ## The Danger of Statistics

My large Government employer conducts a survey every few years to see how management-officer relationships are going. The survey results get broken down in a variety of ways and one of the ways it was broken down enables my business area to be compared with the equivalent business area in another region as well as the organisation as a whole. The data was also compiled to assess male results against female results.

My Manager(female) has looked at the data and sees that men are more satisfied in our workplace than the women are. She also points out that the disparity between male and female satisfaction in our business area is much more significant than the organisational average and our equivalent business area in another region.

She has concluded that we have a gender-based problem.

When I look at our business area, I see things that make me disagree with her conclusion (and the logic she used to reach the conclusion)

Facts:

Our business area has a small population
Women are in a minority
One team is all female, except for its team leader
The females in that one team represent 50% of all females in the business area
I have a strong suspicion that the Team leader in that team is a T whereas the female team members are Fs
We are an engineering organisation and women in that team are not engineers

I feel that if the data could be compiled to compare the workplace satisfaction of engineers vs non-engineers, or Ts vs Fs, we would find greater disparity than we have found by comparing Male vs Female

I'm seriously contemplating telling my manager to put the statistics in the bin and that we have a small enough population to simply have each team leader go up to each team member individually and ask "how can we do better?".

What do others think?

2. Yeah.

3. Originally Posted by mikamickmac
My large Government employer conducts a survey every few years to see how management-officer relationships are going. The survey results get broken down in a variety of ways and one of the ways it was broken down enables my business area to be compared with the equivalent business area in another region as well as the organisation as a whole. The data was also compiled to assess male results against female results.

My Manager(female) has looked at the data and sees that men are more satisfied in our workplace than the women are. She also points out that the disparity between male and female satisfaction in our business area is much more significant than the organisational average and our equivalent business area in another region.

She has concluded that we have a gender-based problem.

When I look at our business area, I see things that make me disagree with her conclusion (and the logic she used to reach the conclusion)

Facts:

Our business area has a small population
Women are in a minority
One team is all female, except for its team leader
The females in that one team represent 50% of all females in the business area
I have a strong suspicion that the Team leader in that team is a T whereas the female team members are Fs
We are an engineering organisation and women in that team are not engineers

I feel that if the data could be compiled to compare the workplace satisfaction of engineers vs non-engineers, or Ts vs Fs, we would find greater disparity than we have found by comparing Male vs Female

I'm seriously contemplating telling my manager to put the statistics in the bin and that we have a small enough population to simply have each team leader go up to each team member individually and ask "how can we do better?".

What do others think?
I agree with you and would be skeptical as well. What do you think the consequences of giving your suggestion would be?

4. You are right. Approach it that way.

With some Fe applied, it should go over reall well.

5. Numbers don't lie, people do.

Anytime a stat is presented it's important to consider all the details of it's compilation, who is reporting it, who owns those people and what can be gained from the knowledge.

Ofcourse gaining something doesn't rule out the truth, but it should definately be looked at.

6. The only problem with statistics is people using it incorrectly.

So yeah, I agree.

7. Well, I gave my Manager my thoughts and it went down well. She particularly appreciated my view re the all-female team. Something I neglected to point out in the OP, but didn't in my email to her, was that this team of all-female Fs led by a Male T that fullfils a non-engineering function in an engineering organisation has, within the past 12 mths, been told that the organisation no longer wishes to participate in the arena that that team participates in. So add job security for a team that just happens to be female as another valid reason for skew.

8. if you ran the stats against all of the variables properly (i.e. jobs held by those individuals, length of employment, section in which they work... obvious things of that sort) you can illustrate the truth in NUMBERS!

people always fail to take into account all of the proper descriptive details when analyzing statistics... they just look at the broad numbers without questioning them because most people suck at numbers

this is why it's so easy to lie with statistics

9. Employee surveys are a method for strategic or upper management to avoid attention to their own mistakes imo. Or it shows that they dont have enough competency.

Employee satisfaction is irrelevant if you pay enough or run a transparent company. Especially in small or medium-sized companies it's good when the employee identify with the company. The last company I worked for, send a monthly newsletter to every employee reporting about contracts and giving them a chance to participate in management decisions. This was a really great tool and enabled me doing an internship, to get invited to a board session because I had a good idea. This is what motivat6es you, when you A see potential for growth of yourself, namely that your ideas are intresting or that you get more responsibility in the company and B that your payment is awesome.

Too many companies start to look at internal structures when the business is running bad, instead of looking at the products or services and reinventing them or inventing new to better their market position. Internal structures aint really too intresting in small businesses, they become intresting in huge automated facilities but in small- and medium-sized the social happiness and a kind of family structure between the people should always prevail before efficiency. Efficiency comes automatically when people are motivated.

If people really have problems they can in Germany go to the companies Union, which directly disputes stuff with the management. Every company needs a Union if it reaches a certain size and this is controlled by the branchs regional Union, so even if your companies Union was korrupt, there's a higher instance controlling the lower ones. This works very good often too good and often is better for employees than employers.

10. Originally Posted by mikamickmac
Well, I gave my Manager my thoughts and it went down well. She particularly appreciated my view re the all-female team. Something I neglected to point out in the OP, but didn't in my email to her, was that this team of all-female Fs led by a Male T that fullfils a non-engineering function in an engineering organisation has, within the past 12 mths, been told that the organisation no longer wishes to participate in the arena that that team participates in. So add job security for a team that just happens to be female as another valid reason for skew.
Yeah, I think all of the points you mentioned blatantly needed to be factored into any analysis of the data; the context of the data had apparently been ignored in the first analysis.

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•