Week 1: Introduction to Scientific Research Flashcards
Necessary, sufficient, and contributing conditions
A necessary condition is a condition that must be present for an event to occur. A sufficient condition is a condition or set of conditions that will produce the event. A necessary condition must be there, but it alone does not provide sufficient cause for the occurrence of the event.
Example: are guns a cause of crime?
- Some say guns don’t cause crime, criminals do, you also don’t need a gum to commit crime, you can use other tools and weapons.
- The claim is guns are not sufficent condition for crime because lots of people have them and they don’t cause any
- Also not a necessary cause of crime, because crimes can be committed without guns
Does this disprove the argument that guns cause crime?
No, because the question is not if guns are the only cause of crime but a cause of crime
Necessary and sufficent conditions are important in the debate but very often you miss the broader picture where the question was a much broader one
5 Logical Fallacies:
1: False dilemma/dichtomoy, equivalence - Either medicine can explain how she healed or its a miracle. Medicine cannot explain how she healed so it’s a miracle.
Either the government created jobs or it did not work. What do we mean by work? We can define it differently.
The problem is that the framework itself sets the discussion which causes an either/or.
Occurs when a limited number of options are incorrectly presented as being mutually exclusive to one another or as being the only options that exist, in a situation where that isn’t the case.
2: Hasty Generalization - All billionaires I’ve met are eccentrics. Bill Gates is a billionaire. Therefore Bill Gates is eccentric.
How many billionaires have you met to make a generalization on all of them?
The evidence is insufficient to make that claim
The evidence itself is anecdotal
3: suppressing relevant evidence:
Wind turbines kill an astonishing 500,000 birds a year in the US
Imagine there is a large margin or error so +- 200,000 birds a year, does this mean anything?
500,000 birds sounds like a lot of birds, but is that a lot in the grand scheme of things? Here we need a reference point on the total amount of birds. We can also find information of other causes of death for comaprasion. If we find another cause that results in 500 million deaths per bird, does the wind turbines still matter? But we can also deepen our understanding, finding how these deaths happen, what bird species are often caught, which wind turbines, and can lower deaths like this.
4: Neglecting probability:
Are you more scared of driving a car or flying in an airplane?
Probability of dying in a plane crash is much smaller and even though you know the evidence differential you still feel more scarred which borders on cognitive biases.
5: Slippery slope:
We can’t have pollution control laws, because if we have such laws, then we will have laws on more and more issues, and we will end up with a totalitarian.
You can have logical premises alone, but the links of each premise together is flimsy and not certain and you end up with a illogical or unconvincing conclusion
4 Cognitive Biases:
1: Confirmation bias: We are more likely when we have an idea about a question to look for evidence that confirms what we think is the answer.
2: Observation selection bias: the way you look at certain pieces of evidence are linked to your confirmation bias
3: Projection bias: when you put yourself in the position of someone else and basing your reasoning on how you would of reacted in such a situation
4: Current moment bias: people tend to favour more things that will impact the current the moment rather than things that will happen down the road, this is a trouble that policy-makers deal with.
A systematic approach to analyzing arguments
he systematic way to treat arguments, we must ask the following questions:
What are the premises?
What is the evidence supporting the premises?
Where does the evidence come from?
Are the links and reasoning logical, coherent and measured?
Are there definitional ambitious in the wording?
Do you agree with the definitions, assumptions and methodology?
Strengths of quantitative analysis
Describe and simplifies information with numbers
Generalize from the information
Infer from the information to better predict
Precision
Measures are more objective - something similar across anyone answering the question regardless of who it is
Comparisons are easier
Communicate information more easily
Manipulate and communicate numerous details
Weaknesses of quantitative analysis
Simplification means incomplete information: What is being left out?
Measurement issues: are we really measuring what we want to describe
What is the validity, reliability and precision of the measure?
Nominal variable
no mathematical relationship to one another, things such as religion, hair colour, or things that just cannot be ranked but you can still code
Ordinal variable
Variables that can be ranked but the exact distance or difference between the values is not the same across all values and hence is not fully known. For example, questions that ask strongly disagree, disagree, neutral, agree, strongly agree
Interval variable
can be organized into order and can be added or subtracted but not multiplied or divided because the zero points is statistically arbitrary. The temperature of zero does not be the absence of temperature.
Ratio variable
Can be ranked into an order, can measure the distance between them, can add, subtract, multiply, and divide and the zero means something, it equals the absence of the thing you’re measuring.
Mutual exclusivity and collectively exhaustive
○ Mutually exclusive: no case can be measured as having two different values, the definition of categories needs to be so cases cannot fall into two of them
○ Collectively exhaustive: There must be a value that is ascribed to any potential case that we may encounter - it needs to be impossible to have a case that would not have a value
