Entering Confidence Intervals In Numbers

Question

Entering Confidence Intervals In Numbers

Hey guys,

It's that time of year again for college students, and I need some help. I have a big lab report due and I'm struggling with the numbers program. I used a separate program to generate the data I need for the lab, and I've entered it into a scatter plot and graphed a regression. I would like to enter confidence intervals specific to each data point but numbers treats every point in my graph as a unit, not an independent sample. Basically I'm saying that if I try to enter the CI's for one point, numbers will produce the same error bars for every point on the graph. This is, obviously, not ideal nor useful. I've tried doing it the laborious way and entering each point individually in the Y-value data entry bar, but no use. This is infuriating and is very well threatening my grade at the moment. Any help would be awesome.

MacBook Pro (Retina, 15-inch, Early 2013), OS X El Capitan (10.11.4)

Posted on Apr 25, 2016 5:27 PM

Reply

Answer 1

Apr 25, 2016 7:07 PM in response to SGIII

I hope all the good things in the world happen to you and everyone close to you. I hope each day you wake and find life has begun anew, somehow seeming even brighter than the day before. I hope you never go hungry or destitute, and that you may always have a place that makes you feel that you truly belong. I hope that you overcome every obstacle you encounter in your life. I hope that you find the answers to all the questions you hold within and towards the universe. I hope that you live to a ripe old age with every decade in memory filled with laughter and comfort. I hope your cereal each morning doesn't get soggy. I hope that every new haircut and pair of shoes you buy look fly. I hope you find a 20 on the sidewalk when you leave your house next. I hope that new season of that show you really like is even better than anyone thinks it's going to be. You sir, have saved me and I will never forget this. I might even tell my kids about you one day. And I hope my kids tell your kids about this someday. Then so on and so forth until my descendants remember you ask a deity among men and your name, my dear friend SGIII, is never be forgotten.

Reply

Answer 2

SGIII

Level 8

37,562 points

Apr 25, 2016 6:12 PM in response to College Stress

You need to have columns (or rows) in your table that contain calculated values for the error bars. See this explanation, which uses the STDEV function. The error bars part starts at 4:45 or so.

SG

Reply

Answer 3

SGIII

Level 8

37,562 points

Apr 25, 2016 6:55 PM in response to College Stress

The linked example (the author says the data isn't great but is good for illustration purposes) might work out to something like this:

SG

Reply

Answer 4

SGIII

Level 8

37,562 points

Apr 25, 2016 8:11 PM in response to College Stress

Glad to hear the example helped. I know how to get Numbers to do certain general things in a mechanical sort of way, but please don't depend on my answer to give you the correct results for your data! I don't really know what I'm talking about when it comes to that, and that part is of course up to you.🙂 Thanks for the gold star and the kind feedback (it's the nicest I've ever seen here) and good luck with your studies!

Don't hesitate to post back if you run into further problems as I know there are real scientists and mathematics experts in other time zones who can help.

SG

Reply

Answer 5

Yellowbox

Level 7

21,945 points

Apr 26, 2016 1:21 AM in response to College Stress

Hi College Stress,

I would like to enter confidence intervals specific to each data point

Great video link from SG.

Please ensure that you are using the appropriate statistic. STDEV is a start, but:

https://en.wikipedia.org/wiki/Error_bar

Error bars often represent one standard deviation of uncertainty, one standard error, or a certain confidence interval (e.g., a 95% interval). These quantities are not the same and so the measure selected should be stated explicitly in the graph or supporting text.

I am not a biometrician. Perhaps you should seek advice from your professors. Good luck with your studies.

Regards,

Ian.

Adding to SG's reply, your feedback is the nicest I have ever seen 🙂. Ian.

Reply