Help on data correctness for ANOVA

#1
Hello friends,

My aim is to study the nature of length of stay of vechicles in a parking lot. A day is split to 3 sessions namely Morning - 10am to 2pm, Afternoon - 2 pm to 6 pm, Evening 6pm to 10pm. Given the In time and Out time of a sample size of 9000 vechicles, I need to statistically prove whether there is a significant difference between the average length of stay between the 3 sessions of a day.

I tabulated the time (in minutes) each vechicle spends in the parking lot for each session of the day and my data looked like this below.

1595062134555.png

Now to perform one-way ANOVA, I have a confusion whether I have to consider the records without zeros (image marked 1) or with zeros (image marked 2). Kindly requesting

1595062250701.png

OR

1595062341932.png
 

katxt

Active Member
#2
How about classifying each vehicle as Morning, Afternoon or Evening and use the total time for each car. So the first car is Morning 214. Then a one way anova on the three groups. (No zeros.)
 
#3
@katxt I initially did that. But it wasnt looking correct.
Say intime is at 1.30 pm and out time 8 pm.. if we use intime alone to categorize like you said, then the entire stay will be categorized as Morning - ‘10am-2pm’ bin.. right? But that is not the actual case here...
 
Last edited: