Virtual Science Fair

Step 1 - Decide on your question 

Step 2 - Decide what data you are going to collect to answer your question 

Step 3 - Identify your variables 

Step 4 - Identify an existing data set or collect your data 

Step 5 - Analyze the results 

Step 6 - Present!  

 

Step 1 : Decide on a question to answer (your objective).  

 

First step is to decide what you will study in the form of a research question.  

 

  1. Think of a research question. Look around your home, neighborhood, or school and see if there is a question that interests you. For example, you can ask, “Does exercise improve my mental health? Does it make me happier?” You want to frame your question as independent or dependent variables. Here you are looking to see if exercise (your independent variable) makes you happy (your dependent variable).   

  1. Create a  hypothesis. What is a hypothesis? - It is a guess to the answer of your research question. You can guess either that your independent variable positively impacts your dependent variable or negatively impacts your dependent variable. For instance, you can guess that exercise will make you happier.  

 

Step 2: How do we answer this question? What data do we need?  

 

Next, turn your research question into an experiment, survey or other appropriate data collection method with formal hypotheses.  

  1. How to define your focus population: Be specific For example, if you want to see how late students stay up, what students are you interested in? Your class, the entire school, all students in your grade in the country? Honing this target population will help focus the project. 

  1. What kind of data do you need?  In other words, is your data categorical (either ordered or unordered categories), continuous, or binary?   

  1. How much data do you need to collect to answer your question? 

  1. It is important to remember that any answer to your hypothesis matters.  In other words, even if there is no relationship between the independent and dependent variables, that is useful information. 

 

Step 3: Identify your variables 

 

  1. What are your independent and dependent variables? 

  1. Are there other variables that could impact the relationship between the independent and dependent variables (i.e., are there any confounders?)? 

 

Step 4: Identify an existing data set or collect your data 

 

  1. How will you collect the data in a systematic and unbiased fashion?   

  1. How many trials or repetitions of the experiment would be sufficient? 

  1. What supplies will you need to collect the data? 

  1. Where and when will you need to collect the data? 

  1. Is there an existing data set that includes what you need to answer your question? 

 

Step 5: Analyze the results 

 

  1. Describe your data 

  1. For continuous variables, it is best to report the median, interquartile range, mean and standard deviation.  

  1. For categorical variables (including ordered, unordered, and binary), it is best to report frequencies (%).  

  1. Be sure to include those descriptive statistics (#1 and 2 of this section) in a table and include all variables that are relevant to your experiment. 

  1. How can you use visualizations to further explore your variables? 

  1. For instance, a box plot or histogram can be useful for looking at the distribution of a continuous variable, and a bar graph can display categorical data.  

  1. If there are any outliers in your data, evaluate whether or not those should be excluded from the analysis.  

  1. To take your project a step further, consider testing your research/project hypothesis using inferential statistics or statistical modeling (optional) 

  1. Note: Some of the time descriptive statistics will be good for the project 

 

Step 6: Present/Discuss Results 

 

  1. Begin organizing your project: 

  1. Introduction: Think about the hypothesis of your project and research it. Compile a summary of the insights you find from this process, citing your sources. 

  1. Methods: Include a data collection form or table. Consider, what else should be included? What methods would improve your project? 

  1. Results: You can’t write this section until you’ve collected your data and completed your analysis, but creating blank data tables and graphs as templates to come back to when you have results will help you think through the project’s design and execution. 

  1. Once data collection and the analysis are complete: 

  1. Enter data into the tables and explore it in graphs. 

  1. Complete the assessment and conclusion sections of the report. 

  1. Some ideas to keep in mind as you discuss your results: 

  1. Remember to use precise language and explain in a way that those not familiar with the subject can still understand. 

  1. If you did inferential statistics or statistical modeling, are your results statistically significant?  What evidence supports your answer?  

  1. It is important to note that if an association between variables is observed, this does not mean that one of those variables causes the other.  In other words, association does not equal causation.  

  1. Experiments could be biased for a variety of reasons (measurement error, selection bias, etc.).  Therefore, the results are not fact and only provide evidence of one answer to the hypothesis.  Experiments and analyses often are repeated by others to provide more evidence of a particular theory and to extend that theory.