browser icon
You are using an insecure version of your web browser. Please update your browser!
Using an outdated browser makes your computer unsafe. For a safer, faster, more enjoyable user experience, please update your browser today or try a newer browser.

Graphing 101

Once you have your data, you will need to present it to your teacher and science fair judges. In a science publication, you would choose between a table and a graph, but for the science fair project it is acceptable, and even encouraged, to showcase the data in both forms.

Type of graphs: Your first choice is to determine which type of graph would best communicate your findings. Your basic choices are bar graph, line graph, pie chart, or scatter plot.

BAR GRAPH – This is the most common type for science fair projects. You may select a bar graph when your independent variable is qualitative (categories) or quantitative (numbers).

An example of a bar graph

Look closely and make sure your bar graph has all the highlighted parts:

Independent variable may be qualitative or quantitative

LINE GRAPH – This is the second most common, but frequently used incorrectly, so be careful here. You should only select a line graph if your independent variable is quantitative (numbers) and you hypothesized that the changes in the independent variable would result in changes in the dependent one. For example, line graphs are great for showing changes in the dependent variable over time or distance along a transect.

An example of a line graph

Again double check the axes:

Independent variable MUST be quantitative

PIE CHART – Pie charts are good for projects that have qualitative independent variables and have generated data that can be expressed as percentages of the total. For example, if your data were counts (i.e. the number of times something happened), then this might be your best choice to compare different treatments.

An example of a pie chart

SCATTER PLOT – If the purpose is to see if the variables are related (common in environmental projects), but there was not a clear choice for independent and dependent variables (for example wind speed and water temperature), then a scatter plot would be your best choice. This option typically requires much more data than the others to observe a trend.

Scatter plots are also called X-Y plots


38 Responses to Graphing 101

  1. Chris Harrris

    This is Really Helpful

  2. Annette

    My daughter did an experiment to see which long stemmed flower would last the longest. What would be a good graph or is it possible to graph the results?

    • Dr. Maille Lyons

      Bar graph: flower type on x-axis; time on y-axis. Hopefully you had at least 3 flowers (e.g. 3 roses) for each flower type. Present photos if you have them to show when you decided the flower was “done”

  3. Melissa

    I’m doing an experiment on: How does the amount of water affect the growth of plants? I’m conflicted in which type(s) of graph I should use.

    • Dr. Maille Lyons

      You could do a bar graph, for each different volume of water used (x-axis) with the final growth (height? weight? number of new leaves? etc.) on the y-axis.

      You could also do a line graph with time on x-axis and growth on y-axis, using a different colored line for each volume used, if you measured growth over several time points.

  4. Desiree & Alaira

    We LOVE this idea…..our project is due in TWO MORE DAYZZZZ!!! and we didn’t even start on our projects yet because we didn’t know which one to pick…..and we have NEVER done a science project in our lives! the last part we needed was a chart to go with what our experiment was about….and all we wanted 2 say was thnk you! with love from Beaufort Middle School (South Carolina)

  5. Sophia Parker

    My son is doing a Science Fair Project on bacteria I have 8 different samples and 3wks of work what chart do you think I may need?

    • Dr. Maille Lyons

      What are the 8 samples?
      What did you measure?
      What do you mean by 3 weeks of work (i.e. what did he actually do?)

      Most likely a bar graph (based on 8 different types of samples), but potentially a line graph because time (3 weeks) is quantitative.

  6. Diego Morales

    I am doing a science fair project on if music can improve taking free throws in basketball and I don’t know what graph to use

  7. Zoe Moore

    Did a science project on which substances made metal rust. No time involved. All metal rusted in different solutions

    • Dr. Maille Lyons

      Did they all rust the same amount? What was your control (plain water – maybe?) Did the control also rust as much as the others? Do they look different in ANY way? What metal objects did you use? What solutions did you test?

      • Zoe Moore

        Thank you. The penny and nail and paper lip all rusted at different time frames. Control was water and it rusted as much as the others. Metal button rusted at different rate. We also had the nail in mud and it rusted. The penny and button had white and black on it. The nail had orange on it. The paper lip was just black

        • Dr. Maille Lyons

          Your best bet will be photos with a data table for your observations (i.e. what you described each as). There really isn’t anything to graph unless you can somehow estimate the percent coverage of rust.

  8. Aden

    I am doing a science fair project on changing the color of carnations. I am using red, blue and green dye.
    I will try to find which color is absorbed the fastest. LIght, amount of wate, temperature of water and the number of dye drops will all be the same. I dont have a clue as to how to graph this project. Any ideas? Please help!

    • Dr. Maille Lyons

      If you measure the time it takes to first see color and/or the time it takes to get a fully colored flower, you will be able to make a bar graphs for time vs. treatment (each color).

      Alternatively you could photo at the start, then 1 hour, 6 hours, 12 hours, 24 hours, then once each day until you don’t see any further color changes (or any other time periods you think are best). You will then be able to make a line graph with time on the x-axis and percent colored on the y-axis. You will have three lines (one for each color). You can estimate percent coverage from the photos and compare.

  9. Jack

    My science fair project deals with behavior science. I am testing about 10-20 people between the ages of 15-60. I want to know how to make a graph or chart showing how well people remember a short poem. I am documenting how many pauses and mistakes and if they have to start over. I do not know how to show my findings though. Please give me advice. Thank you.

    • Dr. Maille Lyons

      You could group the data by male and female and make a bar graph for number of mistakes (y-axis) vs. male/female.

      You could group the data by age (pick 5 or 10 year categories like: 15-20, 20-25, 25-30, etc.) and do the same as above.

      You could also make a scatter plot graph with age on the x-axis and number of mistakes on the y-axis to see if older trends toward more mistakes.

  10. Anita Rodriguez

    I am doing a science project on which gum is stickier, chewing gum or bubble gum. I measured the tack of the gum by sticking each piece of gum (each piece cut to 2 grams) “sandwiched” between two pieces of foam board. I pulled the foam board apart and measured how far it stretched (stuck together) until it fell apart, at which point i measured the distance in cm. At that point I also recorded the area that was still covered in gum and recorded the percentage of gum still stuck to the foam board. Is there another (or better way) to measure the stickiness of the gum? How would I put the data on to a graph?

    Your help is much appreciated!

    • Dr. Maille Lyons

      Sounds like you have already made the project your own and have quantitative dependent variables which is exactly what you want! Nice job.

      OK – for the graph, you can make at least 2bar graphs with Gum Type on the x-axis: 1 bar for chewing and 1 for bubble; then the y-axis will be Stickiness, measured as the length reached before breaking (in cm) = 1 graph and Stickiness, measured as the amount left stuck behind (in % still stuck) for the 2nd graph.

      If you want to do more work, you could also use your system, rigging a small cup/holder to the second board and see how much weight it can support before breaking (so instead of pulling up, let gravity pull down) – your metric would be the number of pennies (or other small weights) the gum can hold before breaking.

      You could also test more types of gum (e.g. different brands of each type) or if you can somehow get an estimate of sugar content you could compare “stickiness” vs. sugar content.

  11. Kimber

    My child is preparing for a science fair project – wants to know does different types of paint affect the drying time of the paint? The indepentent variable being the different types of paint. We have identified the level of dryness as the dependent variable though we are stuck on how to measure & then graph using quantitative measures since it is more qualitative. We identified a nonstandard measuring tool as a rubber glove with identifying 3 different levels of dryness (adjectives- wet, dry and tacky) by applying finger- touch pressure. We are not convinced this is correct- even though part of the experiment will be to test in certain, set increments of time & record the level of dryness. Would the time increments be the dependent variable instead of the level of dryness? Please offer any thoughts on this and graphing suggestions. I believe the x axis is to show the independent variable (types of paint), but it makes more sense to have the increments of time on the x axis and level of dryness (dependent variable) on y-axis & the graph show the different types of paint. Thoughts?

    • Dr. Maille Lyons

      Every time you “touch” mark the time. Best if you “touch” at regular intervals of time (e.g. every minute, every 5 minutes, etc. whatever you feel will capture differences) – make sure you touch different parts of the flat object to be painted. I recommend painting something long and skinny (e.g. 2 by 4 piece of wood) so that you can systematically touch it and keep track of wet and tacky and dry. Take photos of the rubber glove to show the difference between tacky (some paint) and dry (no paint on glove).

      Then your dependent variable is level of dryness, measured as the time it takes to stop being tacky (quantitative).

      You will be able to make a bar graph with the type of paint on the x-axis and the average time to dry (as measured by touching) on the y-axis (because you should have at least 3 painted objects for each type of paint).

      You will also be able to make a data table with time across top (columns) and type of paint as rows with “wet” or “tacky” or “dry” in each corresponding box. If you touch systematically along a 2×4 then you should also have great matching photos.

      Data could look something like:
      Type 1: wet, wet, wet, tacky, tacky, dry
      Type 2: wet, tacky, tacky, tacky, tacky, dry
      Type 3: wet, wet, wet, wet, tacky, dry,

      You will not be able to make a line graph because “wet/tacky/dry” are not quantitative, but you have outlined a creative way to answer the question.

  12. Dawne Sheffield

    My son is making a crystal radio for a science experiment and is just testing whether it works or not. So it is yes or no results, how should this be graphed. What would the independent and dependent variable be ?

    • Dr. Maille Lyons

      This sounds more like a project and less like an experiment… There do not really appear to be any independent and dependent variables. Perhaps if you add more details – did he ask a question? did he test any particular aspects of the building? did he try different crystals?

  13. Kim Sprouse

    My daughter is doing an experiment on Bubble gum – Which graph would I use? She needs to show which one lost is color the fastest on Trials 1-10 and we used grape, strawberry and sour apple – Thanks!

    • Dr. Maille Lyons

      Bar Graph! Sounds like you measured “time” as your dependent variable, so would go on the y-axis (average of ten trials); and “Type of Gum” or “Gum Flavor” would go on the x-axis, then label each bar accordingly (grape, strawberry or sour apple). I would couple this with a data table to clearly indicate you did 10 trials for each flavor. Or you could make an individual bar graph for each flavor, with time on the y-axis and trial number on the x-axis, showing 10 bars. This would be an indication of how variable your method of determining when the color was lost was. Good Luck.

  14. sarah

    My son is doing an experiment on erosion,he needs to find the results and is wondering if a pie chart would work best,he put a plant and soil in one pan,the other sand,and another soil.Which one should he use?Thanks!

    • Dr. Maille Lyons

      What did he measure? From the details provided, I would generally recommend a bar graph with each “treatment” on the x-axis (i.e. each condition he tested) and whatever was measured on the y-axis (hopefully a number of some sort).

      A pie chart is more for percentages – for example % yes vs. % no to a particular question; Or the % of people that picked a certain color; etc.

  15. Ashley Ford

    My son is doing a science fair project how many pennies does it take to sink a boat made out of aluminum foil what would are graph be

    • Dr. Maille Lyons

      That’s not really a science fair project unless you made several boats with something changed for each one (e.g. size, shape, material made out of). Then you would make a bar graph with type of boat on the x-axis and the number of pennies need to sink on the y-axis.

  16. emily

    what website could we use to make a bar graph online

  17. Rose

    Hi, my son is doing a science fair project and wanted to see what kind of graph should he use. His project is he tested 6 people to guess 4 flavors of kool aid while blindfolded and see who guessed all correctly. Some only guessed 2 right and so on.

    • Dr. Maille Lyons

      He could do a bar graph with each person on the x-axis and the number of correct guess (or percentage correct) on the y-axis

  18. PATRIQUE LARCO

    Hello:
    What kind of graph could I use to display the distribution age by two different treatment groups?

    Spinal Cord Simulation vs. Physical Therapy

    Sample Size
    •Much Bigger •Much Smaller

    Age
    •Higher mean •Lower mean

    Age
    •Bigger range in population age (4 to 72 years old) •Smaller range in population range (11 to 59 years old)

    Sex
    •Close to 40% and 60% distribution •Close to 20% to 80% distribution

    Duration of disorder
    •Higher mean •Lower mean

    Duration of disorder

    •Bigger range in population’s duration •Smaller range in population’s duration

    • Dr. Maille Lyons

      It’s not clear from the question what you measured… But if you have two treatment groups, they would usually go on the x-axis of a bar graph with whatever you measured (a number) on the y-axis to show how they compare. For example, if I had coffee drinkers vs. soda drinkers and I measure average height, I would make a bar graph with 1 bar for coffee drinkers and 1 bar for soda drinkers and then each bar would be the average height (height on y-axis).

      But from your groupings, I am not convinced that is what you have for data…

      If the two treatments were: Spinal Cord Simulation vs. Physical Therapy, then these would go on the x-axis and you would have 1 bar for each. Your y-axis could be age (1 bar each to show which treatment had older/younger subjects).

      Usually if you are interested in age distribution, you would make categories (for example group the data by year-class of 5 or 10 year intervals) and then graph the thing you want to show (maybe the number of subjects that use physical therapy?) for each age class. If you did one color bar for physical therapy and a different color bar for spinal cord simulation then you may be able to show there is a difference in the typical ages of who prefers which treatment…

      Hope that helps! If not, re-phrase you question or be more specific on the data you have.

  19. PATRIQUE LARCO

    Thank you that’s perfect. I really wanted to see what kind of graph(s) could be used to display the distribution of age by treatment group?
    I think you’ve made it clear for me.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>