[MedicalMayhem]

Hi, I'm doing a report for biology and had some trouble drawing the graph on the computer. The problem for me is that the x axis does not increase in fixed steps, so it's not 1, 2, 3 etc... but it's more like 23.2, 24.5, 25.1. But when I put these values into word to plot the values, it treats them as if the gap between the values are the same, whereas I wouldn't mind something like 23, 24, 25, 26 on the x axis with points in between or something. Could someone explain to me how I do this?

Thanks!

[Mad Vlad]

(Original post by MedicalMayhem)
Hi, I'm doing a report for biology and had some trouble drawing the graph on the computer. The problem for me is that the x axis does not increase in fixed steps, so it's not 1, 2, 3 etc... but it's more like 23.2, 24.5, 25.1. But when I put these values into word to plot the values, it treats them as if the gap between the values are the same, whereas I wouldn't mind something like 23, 24, 25, 26 on the x axis with points in between or something. Could someone explain to me how I do this?

Thanks!

Are you using line graph or XY Scatter?

[MedicalMayhem]

(Original post by Mad Vlad)
Are you using line graph or XY Scatter?

A line graph, which preferably would go from 23 to 30 on the x axis, whereas when I tried to draw it, it would start at 0...

[Repressor]

Select and right-click the axis > format axis to get to there.

Then set the min (23?) and max (30?) and the step (Using major(Auto)/minor(1)) you want.

(I've included what I think you want in brackets).

Of course, it may be beneficial to start the graph at 0 for scaling/scientific reasons.

[MedicalMayhem]

(Original post by Repressor)


Select and right-click the axis > format axis to get to there.

Then set the min (23?) and max (30?) and the step (Using major(Auto)/minor(1)) you want.

(I've included what I think you want in brackets).

Of course, it may be beneficial to start the graph at 0 for scaling/scientific reasons.

One problem, I'm guessing you have Excel 2010? I have 2007, I'm not sure if that complicates things I've posted pictures of the format axis popup box on my screen, but I'm not really sure what to do.

So you think it would be more beneficial for me to go from 0-30, even though 0-23 will be 0? My project is on salinity levels affecting activity of dogwhelks, and they don't respond between 0-23, hence I thought it wasn't necessary to include it. But how would I make a graph that is a line graph, on the x axis has a scale from 0-30, and y from 0-180 with a curved line?

Sorry for all these questions. I've uploaded pictures on Imageshack, I really appreciate your help!

http://imageshack.us/g/841/31250570.png/

[Repressor]

Yeah, I think you should use an XY scatter graph as you have two sets of continuous data. This also means the options from the previous posts are enabled (your X axis in the images don't make much sense when compared to the actual data (due to being a line graph, I think)).

Just plot the data as points, then use a trendline to insert a line/curve of best fit.

[MedicalMayhem]

(Original post by Repressor)
Yeah, I think you should use an XY scatter graph as you have two sets of continuous data. This also means the options from the previous posts are enabled (your X axis in the images don't make much sense when compared to the actual data (due to being a line graph, I think)).

Just plot the data as points, then use a trendline to insert a line/curve of best fit.

Thanks, I'm almost done now Just one final question, I've attached a copy of the graph. I've tried a scattergraph without the line, but then the trend lines are either linear, exponential or polynomial, which mine doesn't fit the trend. Is there an option for just a curved line through the points? And then there is the scattergraph option with a curved line, however it goes below the x axis?

What do you think is the best option? As I just want a simple curved line through the points :/



Uploaded with ImageShack.us

[Repressor]

Yeah, Excel trendlines can be hit-and-miss. I'd probably just draw one on with Auto-shapes. A flat line then an r-shaped curve.
Slightly ghetto, but more 'right' than that trendline.

[MedicalMayhem]

(Original post by Repressor)
Yeah, Excel trendlines can be hit-and-miss. I'd probably just draw one on with Auto-shapes. A flat line then an r-shaped curve.
Slightly ghetto, but more 'right' than that trendline.

Thank you very much!

When you say autoshapes, do you mean the one on word or paint?

[MedicalMayhem]

(Original post by Repressor)
Yeah, Excel trendlines can be hit-and-miss. I'd probably just draw one on with Auto-shapes. A flat line then an r-shaped curve.
Slightly ghetto, but more 'right' than that trendline.

One last thing, so I've got this:

Uploaded with ImageShack.us

What do you think? Could it be made better? Do you think it's better to join up all the points, or to make it a curved best fit line. And from the straight line to the curved bit, does it look alright? (Sorry for all these questions, but last thing I promise!)

[Repressor]

The line of best fit is generally accepted as the better thing to do. What do you think about what's happening near the 20-22~ mark on the x axis. Is it a sudden increase? or more of a gentle increase?

All you need to do now is label the axis, give it a title and delete that legend!

[MedicalMayhem]

(Original post by Repressor)
The line of best fit is generally accepted as the better thing to do. What do you think about what's happening near the 20-22~ mark on the x axis. Is it a sudden increase? or more of a gentle increase?

All you need to do now is label the axis, give it a title and delete that legend!

Okay, I think I'll stick with a line of best fit then. Well I read that the dogwhelks are inactive if the salinity levels are below 22, so it would theoretically be a sudden increase. So I suppose mine would be quite accurate then, in that respect?

Ah yes, I was too focused on the graph itself! But the 'legend'?