Reference no: EM132243418

Sample assignment Notes:

Identify three activity or productivity metrics from your workplace. They can be your own responsibilities or those of your whole department.

When I was a software developer for Bayer Corporation, our department tracked the following three productivity metrics:

1) Number of hours worked on each project,

2) Number of production jobs that failed, and

3) Reasons that production jobs failed.

Classify the data you would need to collect to measure your results. For each metric, state whether the data would be qualitative or quantitative. If quantitative, state whether the data would be discreet or continuous.

The number of hours worked on each project was recorded (obviously) as a numeric value. That makes it an example of quantitative data. Since we could report fractional hours, the data could assume any positive value. Hence, this is an example of continuous, quantitative data.

The number of production jobs that failed was also recorded as a numeric value. So, it's also an example of quantitative data. But since a job either failed or it didn't, the reported values could only be whole numbers. Since the data can assume only particular values, this is an example of discrete, quantitative data.

The reasons that production jobs failed were recorded as alpha-codes. After determining the cause, we selected the most appropriate reason code from a predefined list. Since the codes are non-numeric, this is an example of qualitative data.

Specify the Level of Measurement of the data collected.

Since metric 1, the number of hours worked on each project, allows us to calculate both precise differences between values and ratios of values, this is the Ratio Level of Measurement.

Since metric 2, the number of production jobs that failed, allows us to calculate both precise differences between values and ratios of values, this is the Ratio Level of Measurement.

Since metric 3, the reasons that production jobs fail, is simply descriptive with no sense of ranking or order, this is the Nominal Level of Measurement.

Pick one of the metrics you identified in step one and discuss how you did, or could, measure the data.

For the first metric, the number of hours worked on each project, each person was responsible for recording their own time. We used a time reporting system to report the information, entering number of hours and project codes into an online form. This was always done after the fact, sometimes days or weeks later.

Identify any potential sources of error in the measurement process. Classify each source of error as either random or systematic.

I can identify at least three potential sources of measurement error.

First, some people didn't take good notes and simply entered the information from their (vague) memories. Some would over-estimate and others under-estimate their time for any given project. Since the resulting errors do not consistently skew the results towards certain values, this is a RANDOM ERROR in the measurement process.

Second, not all activities could be clearly aligned with a single project. Certain tasks could benefit several projects, and there were no clear guidelines on this time should be accounted for. Some people assigned all of the time to one project arbitrarily while others pro-rated their time across multiple projects. Some projects would be over reported (as when all time allocated to it) and other under reported (when no time charged to it at all even though some benefit was received.) Since this resulting errors do not consistently skew the results towards certain values, this is a RANDOM ERROR in the measurement process.

Third, programmers are expected to be at 100% utilization. That is, all of their time was expected to be charged to projects. But the reality is that we all had down time. Sometimes we were waiting on something or someone else, or we didn't have all the tools we needed. Since no one wants to appear as under-productive, people sometimes charged time to projects even when they are not working on them. This is a SYSTEMATIC ERROR because the “peer pressure” leads to project times that are over reported. In other words, the measurement process skewed the results towards certain, in this case inflated, values.

Identify one way to reduce or eliminate the measurement errors.

Here are some possible ways to reduce or eliminate the measurement errors I identified.

Management could set a policy that time must reported by the end of every day. This would encourage us to record information while it is fresher in mind.

The company could investigate ways to capture programmer activity automatically without requiring us to manually report it. For example, the system could capture time logged into/logged out of systems, the saving of files, etc. System recorded values are more accurate than after the fact recollections.

Defining standard ways to report activities that benefit multiple projects would make the reported results more consistent. That's because you're taking away opportunity for varying, individual interpretations.

Finally, Management could change the practice/culture that expects 100% utilization. A more realistic, tolerant view frees people to more accurately report their time.

Assignment:

Employers often define metrics to measure productivity or effectiveness in the workplace. Examples include the number of patient appointments, total expenses, customer wait time, etc. Answer the following questions using complete sentences based on your current place of employment or a previous employer. Look at the sample assignment to get an idea of the expectations for your responses.

Identify three activity or productivity metrics from your workplace. They can be your own responsibilities or those of your whole department.

Classify the data you would need to collect to measure your results. For each metric, state whether the data would be qualitative or quantitative. If quantitative, state whether the data would be discreet or continuous.

Specify the Level of Measurement of the data collected.

Pick one of the metrics you identified in step one and discuss how you did, or could, measure the data.

Identify any potential sources of error in the measurement process. Classify each source of error as either random or systematic.

Identify one way to reduce or eliminate the measurement errors.