Calculation Based: Chi-Squared Tests

Chi-Squared Tests Formula                                     

Critical Value: the number on the table; figure out your degrees of freedom, and ALWAYS use 0.05

  • This means that you are 95% sure of it being accurate (if you pick 0.05)

Degrees of freedom: number of “choices” that you could possibly have minus 1

If the chi squared value is higher than the critical value on the table, then you reject the null hypothesis. If it’s lower than the critical value, then you accept it


Bar Graphs

  • Units: when looking at graph always figure out the units
  • Margin of Errors:

                there’s a statistical difference between two sets of data if their error bars do NOT overlap.

                When question includes margins of error, always look to see if they overlap

Steps to Drawing Graph

  1. Name the graph: Dependent variable (y-axis) vs Independent variable (x-axis)
  2. Label x-axis: will be the independent variable (units!)
  3. Label y-axis: will be the dependent variable
  4. If included, label the margins of error/standard deviation

Scientific Method Vocab

  1. Null Hypothesis: states that here is no statistical difference or correlation (observed diff is by chance alone)
  • The null hypothesis can be rejected if test group is different from the control group
  • Says that there is a similarity between two explanations or things
    • Key words: resemble, like, similar
  1. Alternative Hypothesis (of a test) says there is an effect/relationship
  • Also says that there is another explanation; difference between two things
  1. Dependent Variable: The thing you are measuring
  • Changes in response to the independent variable
  1. Independent variable: the thing that you are changing/testing in an experiment
  2. Positive Control: is not exposed to experimental treatment but is expected to have effect (Ex: positive control for movement, growth, be active)
  3. Negative Control: not exposed to experimental treatment or expected to have an effect
  • Ex: denatured enzyme, inactive gene….
  1. Control: group that does not receive the independent variable (what you are testing)
  • Control group will have everything the same as experimental group, but will not receive independent variable
    • So you know that any change/effect is bcuz of independent variable/thing you are testing
  1. Experimental Control:
  • A control can also be a benchmark which is used to determine something
    • Ex: (thing) is a control for (process)
  • Justify:
    • “All variables must be held constant”
      • Since control has not experienced the variable you are testing, any effect observed in both groups cannot be attributed to independent variable
    • Know that any response/change is only bcuz of independent variable
  1. Environmental Factors: vary them
  • Ex: vary light, temp, duration of experiment, location
  • Justify: know that any change is only because of independent variable


  • Give short answers!! No need for long explanations
    • Think in bullet points
  • When asking you to compare, explicitly compare group __ with group __
  • If question asks to justify including variable, probs be positive or negative control for something

Describing a trend based on a graph/table

  • Point out specific parts of the graph (ie. groups with 6 mm of x had the highest rate of y)

“Identify” a certain number of things


Use specific words from the question


“Calculate total efficiency”

  • If you have ___ amount of energy, but you only use ____ amount, then the amount that you actually used was the efficient part

Divide the amount that you ACTUALLY used by the amount available

Most Missed

  • When question involves phenotypes changing, answer will probs include either directional, disruptive, or stabilizing selection
  • Graph Questions:
    • When describing trend, specifically point out when the line increased/decreased, and the rate it did so (rapidly or gradually)
  • Explain: only need to give one answer
    • Don’t need to explain the explanation; just say ___ because

Predict: only need to predict, no explanation