Z-score application

#1
Good Afternoon,
I am a novice user of statistics and have a question regarding an approach to a problem I have. I manage a small group in an R&D environment. We are assigned "tasks" which are independent of each other, are generally related, but do not require the same amount of time/effort to complete, i.e. some tasks are harder than others. I have access to monthly, weekly, yearly data for times, specific tasks or the tasks as a whole, etc. What I would like to do is analyze whether we are getting better at accomplishing tasks or not. I tried to apply a rudimentary z-score month over month on hours per task ratio. However, when thinking about it, I am really trying to normalize and compare the means for each month rather than the hours per task for a specific task month over month. Is this even possible? Many of my colleges say because the tasks are differing difficulty there is no way to compare tasks month over month. Please let me know how to solve this problem.
 

hlsmith

Not a robit
#2
If you want to know if you are getting better, why not just make a time series plot. You could also conduct a simple linear regression. Outcome would be time and predictor would be month number, jan 2018 = 1, dec = 12, jan 2019 = 13,...,last month. Neither of these are rigorous, but will provide some information.
 
#3
Thank you for your quick reply, but I am not sure I understand. Since we track time per task I can plot total time or average time per task or per month/week/year. However, "time" remains relatively constant since each person works about 40hrs per week. So if I tracked time vs month I think it would be a horizontal line, slope 0. Since time is "relatively" constant but the number of tasks is changing, i.e. difficulty and/or we are getting better, I want to know if the same time is more effective due to some change we make.
 
#5
Yes, I can plot the number of tasks completed, started, or open for any given time period, month, day, year...Thank you for your help.