Investigating Bioluminescence in Dinoflagellates

[Kat_Marie]

I’ve finished gathering all data from my cultures and now I need to display my data in some sort of diagrams, but the problem is I don’t know which kind to use or how to do it exactly. Would I average my data together or use them the way they are? I’m very confused about this. I’ve never used graphs and charts before. I tested each of the 3 sets of cultures twice a day for 7 days. Any help would be grealtly appreciated.

[SciB]

Hi Kat,

Congratulations on finishing your experiments. I hope you got some interesting results.

The Procedure for the bioluminescence project has suggestions for analyzing and displaying your data:

https://www.sciencebuddies.org/science- ... #procedure

Analyzing Your Data

Graph your data.
Graph the time on the x-axis and the brightness score on the y-axis.
Suggestion: Divide the x-axis into 24 hours, with noon in the middle. Mark the period of darkness for the light/dark cycle with a black rectangle under the x-axis. Add bars for the bioluminescent scores to the graph at the appropriate times. Use different colors for the all-light, all-dark, and light/dark cultures.
How does the varying light exposure affect the bioluminescence of the cells?
Repeat the procedure at least two more times so that you have three sets of data.

The type of graph that you will be doing is called a bar chart because the measurements are shown as bars extending up or down from the x-axis. Scibuddies has helpful information and a checklist on how to analyze and graph your data correctly. Many people use the MS-Office program, Excel, to graph results but there are other programs you can also try:

1. GraphPad Prism
https://www.graphpad.com/scientific-software/prism/

2. Origin
https://www.originlab.com/

3. SigmaPlot
http://www.sigmaplot.co.uk/products/sig ... etails.php

If you have never used a graphing program I would look for a good tutorial video on Youtube. It takes some time and practice to learn the software but it is a skill that you will need, especially if you continue in science. If you have questions let us know.

Sybee

[Kat_Marie]

Will I need to make a graph for each individual day, one containing a whole set of data, and one using all three sets of data?

[SciB]

Hi Kat,

If I read the procedure correctly, it says to make a graph for each day (5 days total) containing the data for each of the three light conditions. The time--8 am, noon, 4 pm and 8 pm--goes on the x-axis and your rating (1-4) of the glow is on the y-axis. How many tubes of dino's did you use for each light condition? The procedure showed 2 tubes of dino's for each exposure, but I hope you did at least three because you will average your readings and the more readings you have the more accurate will be your data.

The procedure says to repeat the experiment twice so you will have a total of three sets of data for each light treatment. You will have to make one graph for each day of five days for each experiment, so that will be 15 graphs. But at the end you will average the daily results of the three separate experiments to get a final set of data.

In my opinion, there are two ways you could show the data in final form. One would be with all three light exposures shown for each of the five days at the four separate times (5 graphs). The other would be to have just one light treatment per graph but with the days 1 to 5 on the x-axis showing a glow measurement at a specific time of day (3 graphs). The first type of graph would compare the three light exposure treatments and that is the recommended method. If your data for one of the treatments shows a change over the five days that differs from the other two treatments, you might want to show this in a graph of glow at a specific time on each day for that treatment. If there is no real difference among the three treatments then don't bother to do these graphs.

I hope this is clear. If you still have questions, please post again.

Sybee

[Kat_Marie]

Thank you so much! This helped me a lot. I only checked the dinoflagellates 2 times a day— 6:30 am and 6:30 pm— so that will go on the x-axis. I only used 2 tubes of cultures for each light exposure, so would averaging them work? Also, for the first graph you mentioned for the final data is just like doing the other 15 graphs? Also do I have to make a computer generated graph for every day if I already have all the data written in a table in my lab journal? Could I just create 3 bar graphs for the average for each trial?

[SciB]

You are very welcome! Glad I could help.

1. Yes, do average your two readings per experiment. Three would be better, but two is better than one.

2. If you do the graphs as recommended in the procedure they all will be the same type--bar graphs with the time of day on the x-axis and the glow rating on the y-axis. The final graphs will show the averages of the three experiments for each of the five days.

3. No, you don't need to graph every day's readings for all three experiments, but I would make a table for each light exposure showing the readings on each of the five days so you can see if there is any pattern over time.

4. I think you need to graph the data according to the instructions--final averages of glow readings of dino's from all three light conditions and all three runs on one graph for each of the five days. This allows you to compare the light output for dino's under each light exposure. You should also graph the averages of the 5-day data from each light exposure condition on one graph to see if the light output changed with time of exposure to the test condition.

Sometimes I have to graph my data several times in different ways before I am satisfied that it shows all the changes and difference that I observed. Did you learn about statistics in school yet? If so then you can also include error bars on your graph. The graphing program can generate standard deviation and standard errors for the average of the three experiments and you can put these on the bars to indicate if a difference is real or just chance. If you have questions about this let me know. Using statistics on your data is really the only way to be sure of the results.

Sybee

[Kat_Marie]

I really appreciate all your help! I wouldn't be getting very far without it. I think this might be my last question. So I've created graphs for every day and now I need to do the final graphs. When you first told me how to do them I understood, but I've read all of our messages so many times I've confused myself. So in the last message you sent me you talked about the final graphs under point 4. Instructions wise, are you referring back to the fourth message you sent me? I'm not really sure if that makes sense like I said I've confused myself. I think what I mainly don't understand is what numbers I'm supposed to average together and whether or not there will be multiple final graphs. Also I'm supposed to include measurement units, but for I used the brightness scale of 1-4. Is there any unit I could add? Thank you again for all the help you given me. I can't express how much I appreciate it.

[SciB]

I am glad my help has been helpful! You need a caring guide when you are walking in unfamiliar territory.

I'll try to clear up the confusion without making it worse...

Data for averaging
1. the readings from the two tubes of dino's for each light condition at each time on each day for each of the three experiments [This is your raw data, not the final data that you will graph for your project display board. You can, of course, graph these readings and keep the results in your notebook just to have a record of them.]

2. the averaged readings for each of the three experiments [This will be the averages that you plot on the final graphs for each day. As I suggested, you can also plot the averages for each of the three light condition on a separate graph of glow on the y-axis vs day on the x-axis.

In answer to your last question about units of 'glow'--you don't have any. You will just have to label the y-axis with something like: Relative Bioluminescence Intensity (scale of 1, Least, to 4, Most)

Please don't hesitate to post again if you still have questions. I want you to understand the project because it makes it a lot easier when you have to explain it to someone.

Sybee

[Kat_Marie]

It's getting better I'm starting to understand more. Okay for the first type of final graph I am going to take all six averages from the raw data of all three light exposures for each day of each trial and make a bar graph. So say for Day 1 I'll take all the averages from the raw data of both times the dinos were checked for each of the three trials and I would graph them on one graph. And I'd end up with a total of 7 graphs that way because I tested them for 7 days. For the second type of final graph I would use one light exposure per graph ending up with a total of 3 graphs. I would take both of the averages from the raw data of all 7 days and graph them with Days 1-7 on the x axis and the brightness scale of 1-4 on the y axis. Is any of this correct? To answer one of your earlier questions, no I haven't learned about statistics.

[SciB]

Good job, Kat! I think you got it--one set of final graphs of glow vs time for each day (7) and one set of glow at one time vs day for each lighting condition (3).

OK, so you weren't born knowing statistics but you need to learn some cuz it is really useful. The one thing every scientist looks for on a graph besides the mean (average) value is the error bars. These are lines placed above and below the top of the bar to indicate how close the average value is to the true value. Your graphing program calculates the standard error of the mean (if you tell it to) and puts error bars on your bar chart. Take a look at the Wiki on error bars and you'll see what i am talking about: https://en.wikipedia.org/wiki/Error_bar

The purpose of the error bars is so you can compare the averages and be able to say whether or not they are really different. Sometimes one mean will be higher than another and you might think that it is different but when you look at the error bars you see that both means fall within the range between them and are not statistically different.

Check out this Wiki on Standard Error: https://en.wikipedia.org/wiki/Standard_error
It's a little technical (some math involved!) but will make sense if you read it over and think about it a few times. It is worth it cuz it will make your results a whole lot more meaningful.

I'm glad I was able to help. Post again and let us know how everything turned out.

Sybee

[Kat_Marie]

I can’t say thank you enough! My science fair is Thursday and I’m almlst finished with my IRP all I have to do is create a few more graphs. I’m going to be including you in the acknowledgments section of it and was wondering what name and title you would like for me to use. Thank you so much again! I’ll make sure to tell you of how things go Thursday!

[Kat_Marie]

Also how should I cite this website so that I give proper credit for the doing this project?

[SciB]

You are so welcome, Kat! It makes me happy to know that I contributed to your success.

Just acknowledge Science Buddies as your mentor--no need to put my name. We are supposed to remain sort of anonymous. When you become a great scientist (or whatever) I will have the satisfaction of knowing I helped a little bit!

All the best,

Sybee