Big Data, Plainly Spoken (aka Numbers Rule Your World)

Know your data 40: when they know where you are

It's finally happened. A big company said the quiet part out loud. PwC, the behemoth U.S. consulting company, is using surveillance data to monitor and flag employees for violating its "back to office" policy, according to a news report (link).

The article concerns the U.K. office. There are some possibilities:

• this is a pilot program, eventually to be implemented worldwide
• other locations may already be doing it but quietly
• other locations may have the same policy but the reporter focuses only on the U.K. office

***

What we're learning is that PwC has the surveillance technology in place to do this, and at least one management team has decided to exploit it to encourage compliance to the work in office policy.

According to the article, staff will be sent a report that documents their work locations. If an employee fails to be in the office three times a week as required, the manager will want to "understand" why.

***

This news reminds me of schools that have forced students to install apps that follow them around. See this previous blog.

 

 

Correlation and causation

One of the snippets making its round on display panels in U.S. elevators this week is:

This is the kind of conclusions that have statisticians collectively shaking their heads.

***

The analyst claims that teams that give high proportions of compliments relative to criticism tend to produce higher performance. Thus, teams should focus on being positive.

It's interesting how we interpret the quotation.

The actual sentence really isn't a causal statement. It says high performing teams have a higher ratio of compliment to criticism than lower performing teams.

What's making me turn that into causality?

First, the title of the snippet cues the reader to interpret the sentence causally. "Compliments are key" suggests that it's the positive feedback that causes better performance.

Second, the statement contains two objects: one is an action (giving compliments) while the other is a metric (team performance). Since a metric can't cause an action, it's natural to interpret it as an action causing a matric to change.

Here's a different statement involving two actions: a study finds that workers who lunch at their desks give high proportions of compliments relative to criticism. Now, I don't think one is causing the other.

***

One problem with interpreting associations causally is that we privilege one direction of causality without justification.

Given that the study is almost surely based on observational data, it is not possible for the analyst to disprove the following alternative claim:

Teams that perform well tend to get more positive feedback.

Here, we believe it's the performance that causes compliments, not the other way round.

High team performance is associated with higher proportion of positive feedback but which is the cause and which is the effect?

***

The snippet is sourced to an article in The Atlantic, which references an old research paper (link).

The paper builds a nonlinear dynamics model as shown in this picture:

You can see that the arrow goes one way - from positivity ratio to team performance. Thus, the causal direction is an assumption of the analyst.

I'm not disparaging using causal assumptions - they should be made clear to the audience.