Statistics Saturday: Wage and Hour Statistics Tools

Sampling Error, n. the difference between the
true, and generally unknown, value of a statistic,
such as the mean average salary, of a parent
population and the value that is estimated from a
sample.

Discussion:

Sampling error in commonly presented in national
surveys. For instance, a survey may report that
27%, with a sampling error of +/-3.5%, of the
employees were interested in applying for a given
job.

The sampling error in this example would
mean that it can be expected that had the entire
population been asked the question, between
23.5% and 30.5% of the employees would have
reported that they were interested in the job in
question.

Declaration design principle #3: Ask the right questions in the right way

(from previous posts)
Once the general focus of the survey or declaration statement is determined, it is
important to make sure that the correct questions are ask of the respondent. The
questions that are asked in the survey should be very focused in nature. The
questions should not be leading, but whenever possible they should be closed end in
nature.

Closed end questions, because they yield a choice based response, allow
the results of the declaration statement to be analyzed in a more quantitative manner.
Furthermore, important issues concerning the wage and hour case should be probed
in a couple of different ways to ensure that all angles of the issue are covered
appropriately. Differently worded questions can yield different responses by potential
class member. When appropriate, the declaration statement document should
include follow-up or check questions that help check the validity of the initial
response.

In the unpaid meal break cases study example above it was important to determine if
the fact that they person did not clock out for a meal break meant that the person did
not actually take a meal break. Anecdotal evidence suggests that many of the
pharmacists took uninterrupted meal breaks but simply did not clock out for the
meals.

Declaration design principle #2: Ask as few questions as possible.

(continued from previous post)

Once the random sample of individuals is selected, the declaration statement or
survey questions must be carefully designed. As a general rule the declaration
statement or survey form, and questions asked of the potential class members,
should be as parsimonious as possible. Shorter forms and questions will tend to
increase the amount of effort and thought that each potential class member puts into
completing the survey.

As a general rule only questions that are germane to the issue at hand should be
included. Questions that ask the respondent about information that can be directly
obtained from other sources, such as human resource or time card records, should
be avoided. For instance, questions that ask about the person’s job title, employment
length or other work schedule related issues should generally be avoided because
this information can be obtained from other sources. The additional work required to
match the respondent's information to the survey responses is generally outweighed
by the benefits of a shorter form.

In the off-the-clock case discussed above, the attorneys, with the assistance of
economists and statisticians, devised a detailed declaration statement document of
the grocery store chain’s pharmacists. The declaration statement document had 10
questions that asked the pharmacists about their meal breaks and the frequency in
which those meal breaks were interrupted. There were also questions about the
individual’s knowledge of the company’s policy concerning meal breaks and their time
clock punch activity.

Declaration statement design principle #1: Use a sampling routine to construct

In many wage and hour class actions, asking only a portion of the individuals to fill
out declaration statements or survey is preferable to obtaining surveys and
declaration statements from every member in the potential class. In many instances,
the potential class members are too geographically dispersed to make a total census
count feasible.

For instance, in the grocery story case study (see previous post) the 700 potential members of the class action worked in over 325 geographically dispersed locations throughout the state.
Most stores had at least one pharmacist on duty 24 hours a day. A total census
count of active pharmacists who would be potential class members would involve
administering a survey over a short period of time to employees spread over 2,800
different work shifts (700 members x 4 shifts). Additional effort would have to be
expended to administer the survey to pharmacists that no longer worked at the
company.

Total census counts, even if they are feasible, may not be desirable because
responses may become contaminated as the declaration statements are collected
from potential class members. It is possible that once word of survey or lawsuit gets
out to all the potential class members, later responses may vary from earlier
responses in non-desirable ways. Individuals who complete a declaration statement
in a wage and hour case after they learn of the lawsuit may subsequently go back
and review pay-stubs or other documents and arrive at the estimate of off-the-clock
work in a different manner from individuals who completed statements earlier. In this
instance, the declaration statements of the later potential class members may be
contaminated and should potentially analyzed in a different manner than the ones
completed earlier.

The bottom line is that even if it were possible to survey every person in a potential
class, you may not want to do so. A well designed sample in a wage and hour class
action will help ameliorate many of the problems associated with attempting to take a
complete census count of the entire potential class.

In the grocery store wage and hour case study, declaration statements were taken
from a random sample of 125 pharmacists across the store’s locations within the
state. The pharmacist sample, which used a technique that is called stratified
random sampling, appropriately reflected the stores located in the northern and
southern part of the state and different shifts that the pharmacist worked.

Declaration Statements and Surveys in Wage and Hour Cases: Part I

A Case Study:

In a recent wage and hour class action, the pharmacists at a major west coast
grocery store chain filed a class action lawsuit in state court that alleged that they
were denied the opportunity to take uninterrupted, duty-free meal breaks. The
pharmacists allege that they were routinely required to answer customer’s questions
and even fill prescriptions during their meal break period. In the state in which the
collective action was filed, the pharmacists were as of January 1, 2001 classified as
non-exempt employees. As non-exempt employees they are entitled to one or two
30-minute duty-free meal breaks depending on the amount of time worked. Across
the state, there are about 700 employees spread over about 325 store locations in
the potential class.

In the lawsuit both the extent and commonality of the off-the-clock wage and hour
violation allegations are at issue. Specially, the defendant’s attorneys are looking to
the use of declaration statements administered to the stores pharmacists to
determine what percentage of the time pharmacists believe their meal breaks were
interrupted by work related duties. In addition, the defense attorneys are looking to
determine how representative the named plaintiffs are of the proposed class of
pharmacists. Finally, the defense attorneys want to know to what degree the
allegations of missed meal breaks vary across the proposed class is also at issue in
this case.

While lawsuits involving missed break are more likely to occur in certain states such
as California which have more defined work and employment laws, and less in
others, the general survey and declaration statement design techniques that are used
in a missed meal break analysis are useful in other types of off-the-clock work cases.
Generally, a survey or declaration statement document that can be used to analyze
wage and hour violation allegations needs to have several main characteristics.

Stats Saturday: Wage and hour terms to know

Stratified Sampling, n. a method of statistical
sampling that draws sub-samples from
different sections, or strata, of the overall data
population.

Discussion:

Stratified sampling routines are used in
employment settings when there are important
differences between different groups of
employees that are being surveyed.. For
example, in a survey of off-the-clock work,
workers at different locations and different
supervisors may have different work culture
that make it less likely