Natural Science as an Area of Knowledge

Natural Science as an Area of Knowledge

Hypothesis testing and Induction

I have friends who are virgo and she is shy, therefore I conclude that all virgos are shy

• What is the problem with this conclusion?

• Is this induction or deduction?

• Does science use induction or deduction?

• What can we do to check whether it's true or not?

Are there virgos who are not shy? Are there people with other star signs who are also shy?

Induction = Specific to General

When I ate apple my headache's gone, therefore all people who eats apple everyday will not have headache

People who drank coke on a weekly basis died, therefore all people who drink coke on a weekly basis will die soon!

what's wrong with this? What could be the other possible reasons why those who drank coke weekly died quickly?

These people exercise daily and they're healthy, therefore all people who exercise must be healthy

What's wrong with this generalization? Are there other possible reason why people who exercise are healthy?

Are there people who don't exercise but still healthy? Are the people who exercise but still get sick often?

YES!!!!

Open Ended

Are there other possible reasons why people who exercise are healthy?

People who often exercise are also people who tend to eat healthy and rest properly and that's why they're healthy! It's not (just) the exercise!

Open Ended

Give one example from your own experience where you did a generalization that might actually be wrong

What to do?

• Avoid background assumptions

• remember that there are different hypothesis/explanations for a given set of data (eg. people who exercise are healthy)

Open Ended

Google about the difference between correlation and causation. Give a very brief explanation about the difference (using example)

Make PPT: How certain is scientific knowledge

• problem with observation (optical illusion-we see what we expect to see)

• problem with induction (generalization)

• problem with hypothesis testing and interpreting result (different possible explanation of data)