# Minitab normality test and stats test for date vs time?

#### algarveamy

##### Registered
Hi, total newbie using Minitab and stats, sorry! But I'm trying to use Minitab to get a p and r value for correlation between date vs time (time of sightings over various dates). I need to run a normality test first, but can't get the option of residuals on this data to test whether its para/non parametric to then run the test (also not sure if I need to run Spearmans or Pearsons or other?) I had to convert the time to total minutes to do a graph in excel but Minitab recognises the data to plot the graph as it is, (so not sure if I need to do any formatting on this). Any help hugely appreciated! Thank you.

e.g.
00:35 11-Nov
02:03 11-Nov
01:21 16-Nov
02:10 16-Nov
00:01 17-Nov
02:59 17-Nov
00:01 18-Nov
01:25 24-Nov
etc.

#### Miner

##### Forum Moderator
What is your hypothesis? What do you hope to prove or understand about this data? Do you hope to demonstrate that sightings occur at predictable intervals? Have you considered looking at the time between sightings?

#### Bev D

##### Heretical Statistician
Super Moderator
Do you mean day of the week vs sightings? Date doesn’t seem to make sense unless you multiple years?

#### Bev D

##### Heretical Statistician
Super Moderator
Also your data are not continuous they are categorical. This matters. Forget the math for awhile. I advise you to first plot your data. Date/day on the x axis and sightings on the Y. OR try a bubble chart. Share with us and we can help.

#### Miner

##### Forum Moderator
Also your data are not continuous they are categorical. This matters. Forget the math for awhile. I advise you to first plot your data. Date/day on the x axis and sightings on the Y. OR try a bubble chart. Share with us and we can help.
They could be treated as ordinal as there is an implicit order.

#### Bev D

##### Heretical Statistician
Super Moderator
They could be treated as ordinal as there is an implicit order.
Yes but since they said they were total newbies to Minitab and stats…

#### algarveamy

##### Registered
Thanks for the replies, I'm basically trying to see if there is a correlation of peak times of activity throughout the year. Camera traps were put out for a 3 month period so far, and then the times recorded for each sighting (from camera trap footage), which is why the dates are random..sometimes got nothing, other days multiple. I can plot a basic scatter graph on minitab using the data which gives me a graph with date on the x-axis running from Nov 22 - Feb 23 for the winter period, and time by the hour up the y-axis from 00:00, 1:00, 2:00......12:00.....22:00, 23:00, 00.00. This shows a concentration of activity from Nov- Dec between 00:00 - 04:00 and visually show's the peak activity. But using the same data I can't run a normality test (which I was told I would need to do first) which will tell me if I need to do a Pearson or Spearman depending if its parametric or non parametric, which should give me a p and r value.
Or with the data I have is this just not possible?
Many thanks!

#### Bev D

##### Heretical Statistician
Super Moderator
Can you send the graph to miner or myself via private message? (I think this works miner?). There is a limit of a min number of posts before you can attach a file…

#### Miner

##### Forum Moderator
5 posts before attaching a file. I don't know about the PM minimum.

You are really dealing with time series data, which can be very complicated to analyze. If you had multiple years of data, you would probably see seasonality during those months. You may be able to show seasonality by time of day with your existing data. You could create a dot plot using time of day as the x-axis and plot the number of observations by time of day.

Correlation does not make any sense at all, and normality is a non-issue.

#### algarveamy

##### Registered
Ok, thank you both and thanks for the info Miner, good to know not to persevere with this with the data I have. I will go back to the drawing board. I'm just starting out with this and have a lot to learn! Thanks again, A

Replies
2
Views
362
Replies
1
Views
2K
Replies
1
Views
1K
R
Replies
1
Views
3K
omezquita
O
T
Replies
5
Views
2K