The verification of a hypothesis is an important step of the scientific method that allows to ascertain the validity of a reasoned conjecture. The typical procedure is to formulate a hypothesis based on the evidence gathered and then verify it through experiments. As you collect more and more data, you will be able to understand if your starting hypothesis is correct; if, on the other hand, it has defects, you can review it and correct it so that it agrees with what emerged from the data collected.
Steps
Part 1 of 3: Ask a Question and Start Testing
Step 1. Start with a question
This question does not constitute your hypothesis; rather it serves to establish an argument and to allow you to start making tests and observations, in order to arrive at formulating the real hypothesis. The question should be about something that can be studied and observed; try to think as if you were preparing a project for a science exhibition.
An example question might be: "Which brand of stain remover can remove stains from fabrics most effectively?"
Step 2. Develop an experiment to answer the question
The best way to test a hypothesis is to create an experiment. A good experiment uses guinea pigs or creates conditions that make it possible to establish whether the hypothesis is apparently valid, through the evaluation of a large variety of data (test results).
In the case of the stain remover experiment, you could proceed like this: stain 4 types of fabric (such as cotton, linen, wool, polyester) with 4 different types of stains each (such as red wine, grass, mud and earth, fat); then, try the four or five best stain remover brands (such as Vanish, Omino Bianco, Bio Shout, Gray) to see which one removes the most stains
Step 3. Start collecting data to answer the question
At this point you should start conducting the actual experiment. In any scientific experiment or hypothesis evaluation, the larger the data set, the more accurate the results will be.
- In the case of the stain remover experiment, you should purchase one pack of each of the top stain remover brands and stain different fabrics with various stains.
- Then test each detergent on each of the stained fabrics (if you live with your parents, you will need to ask for permission to use the washing machine for most of a day).
Part 2 of 3: Formulate and Question the Hypothesis
Step 1. Formulate a working hypothesis
It should consist of a statement about what you believe happens to what you are observing. No starting assumptions are 100% true, but it can be improved by continuing to test. A good guess should be your best guess after conducting several initial experiments.
- For example, if you've done a series of washes to test which stain remover best removes stains from linen, you can use the results to make a guess.
- An example of a good working hypothesis would be: "Vanish is the most effective in removing the most common stains from fabrics".
Step 2. Continue conducting experiments
Once you have formulated a working hypothesis, you should continue testing in order to improve it. You will most likely find that your initial guess attempt is not totally wrong, but that it does not represent the full range of data.
Still following our example, since you have tested only one type of fabric (linen), you will have to repeat the washing experiment with the other 3 types (cotton, wool and polyester) and notice which stain remover eliminates the most effectively. spots
Step 3. Analyze the collected data
Once you have tried all the combinations of fabric, stain and stain remover, you will have 64 different results to evaluate. Analyze all the data your experiment produced (i.e. how effective each type of stain remover was in removing each type of stain from each type of fabric). At this point you can draw general conclusions based on your analysis.
- As tempting as you may be to accept only data that supports your hypothesis, this is neither scientific nor ethical.
- You must take all the data into account and observe any patterns that form, even if it proves that the hypothesis is probably false.
- Keep in mind that getting meaningful results does not necessarily mean that the hypothesis is confirmed, but rather that, based on the data collected, the differences you observed are probably not due to chance.
Part 3 of 3: Review and Correct the Hypothesis
Step 1. Use inductive reasoning
This type of reasoning (also called "bottom-up" thinking) allows you to identify repeated patterns and similarities in the data collected. Be guided by the data in formulating your hypothesis and avoid forcing its interpretation to support the outcome you desire.
For example, if you started your experiment thinking that Vanish was the most effective stain remover but then noticed that it doesn't remove red wine and mud stains well, you probably need to change your working hypothesis
Step 2. Modify the hypothesis
If the data does not support the validity of your assumptions, you can rephrase the hypothesis based on the new information. This is a crucial step of the scientific method: anyone who tests a hypothesis should be able, through inductive reasoning, to correct it based on the results derived from observing a large amount of data.
So if Vanish is ineffective on certain types of stains, your initial working assumption is wrong
Step 3. Come to a definitive hypothesis
Once you have tested, reviewed and tested again, you can draw conclusions regarding your hypothesis. If it needs improvement (or if it's completely wrong), it's time to fix it. A good concluding hypothesis should include what you have learned from observing and analyzing the set of data that emerged from the experiments.
An example of a final and verified hypothesis could be: "Bio Shout is the most effective stain remover for removing different types of common stains from the most used types of fabrics"
Advice
- Deductive (or "top-down") reasoning will not be of much help in testing a scientific hypothesis: it must be based on the experiments you have conducted and the data you have collected.
- Depending on the type of hypothesis you are testing, you may need a control group. If you are testing the effectiveness of a drug, for example, you will need a group of subjects who are on placebo.
- Remember that a null hypothesis (when the control and experimental variables are the same) is different from the alternative hypothesis (when the control and experimental variables are different).
