We have data from an experiment and we want to use subgroups to improve precision. Which subgroups would be most helpful to focus on?

Questions 6-8 refer to the following information. levels.fyi is a company that provides job seekers with compensation data. As people negotiate job offers from companies, they can use levels.fyi to see what other job seekers have gotten. For example, someone with a job offer from Meta to be a product manager can see what Meta product managers typically get in compensation. This service is free. levels.fyi also offers a paid negotiation coaching service which costs $1000. Only about 1% of their users actually pay for this service, and levels would like to increase this share. They ran an experiment offering a discounted price of $500 to a random subset of users. Given their prices and costs, levels.fyi determined that they would want to lower the price for everyone, if the lower price caused an increase in take-up of coaching of at least 0.04 on average. Some data from the experiment are in the table below:   Group Sample size Coaching  take-up average Coaching take-up standard deviation Control ($1000 price) 1,225 0.011 0.10 Treatment ($500 price) 1,225 0.020 0.17

What is the definition of a noisy estimate?

A professional sports team collects data on the sleep habits of its players. They notice that sleeping at least 7 hours the night before a game is associated with performing better in the game the next day. If we are interested in the effect of sleep length on game performance, which of the following would be a confound?

We conduct a randomized experiment and want to assess the quality of our randomization. What kind of variables should we include in our balance test?

Halfway houses are residential facilities for various types of offenders, including which of the following?

Many organization would like to help their employees save for retirement, and they offer various savings programs to do so. However, not all employees sign up for these programs, so organization are always looking for ways to increase savings. One large organization tested out whether  “savings rate prompts” can increase savings. Specifically, over the course of a year, they randomized all new employees to one of two conditions: Control: No communication. Savings rate prompt: Employees received an email saying “employees typically set aside 4% of their income in our savings program.” The outcome metric for the experiment was cumulative contribution to the savings account, six month after joining the program. The table below present some information from the experiment   Group average age (years) % Just high school Average contribution ($)  Standard deviation contribution  Control 22.8 81.6 31.3 25 Savings rate prompt 23.1 81.9 37.0 27  

A work ethic camp is an alternative to prison that teaches job skills and decision making using a cognitive-behavioral approach, followed by intensive supervision probation.

With an aggressive GPS system, the daytime tracking data are temporarily stored and downloaded at night through a landline phone while the offender is sleeping.

Community-based punishments that provide more freedom than prison but less freedom than traditional probation are known as __________ sanctions.