Statistical Fallaices and information Flashcards
Total=
Are just added up as a set of units.
Totals only tell us so much because we have nothing to compare them to.
Are just added up set of units, Totals only tell us so mcuh
Ratio=
Is an expression of a total of one thing as a proportion of another.
The Number of Ss of the Sum of their values on a ‘varable p P per unit of some other class T; a ratio is a total expressed in relative in relative terms.
Helps us to understand the context of a total.
Frequency=
Frequencies tell us how many things in a class have a certain property.
The number of proportion of Ss that have some particular value on a vaiable P
Frequency between class and a property.
Distribution=
But Frequencies just tell us how many things in a class have a property (as opposed to not having property)
Distributions tell us how many things each have properties
The Number or proportion of Ss that each of the values (P,Q,R etc) on some variable.
Distribution of property.
Average (mean)
Averages give us the central value of Ss on a quantitative variable.
Ensure we use the same units of measurements for values.
Median
The median is the central Value in a set of quantitative values.
Averages are sensitive to extreme values in ways that medians are not.
The Gambler’s Fallacy=
(ii)Explain why it leads to unrealible reasoning
It’s when we think statically indeperpend events are not statistically independent and we make incorrect guesses on what’s going to come next based on what’s come before.
Previous factors won’t affect the likelihood of the next one- They are statistically independent.
Past doesn’t influence the future.
Mistaking Statistical significance for clinical significance =
(ii)Explain why it leads to unrealible reasoning
Statistical Significance means that there is a low probability that the results are due to random chance. Sometimes things can be statistically significant but not clinically. This is where the effect of the statistically significant factor is so small, close to nothing- so the Clinical significance is in turn nil.
Conjunction Fallacy=
(ii)Explain why it leads to unrealible reasoning
The conjunction fallacy is thinking that the conjunction of two events is more likely than a single general event.
Likely hood of one Probablility occuring (A) is always more than the likelyhood of two probabalitites occuring (A&B)
Pr (A&B) Vs Pr(A)
Base-Rate Fallacy
(ii)Explain why it leads to unrealible reasoning
Base Rate=
The Likelihood of an event occurring out in the world regardless of what the conditions of a particular situation may be.
The Fallacy occurs when a person misjudges the likelihood of and event because they don’t take into account other relevant base rate information
If we are presented with base-rate information (general information) and information about a specific case, we tend to ignore the general information and focus on the specific case.
File-drawer effect:
Researches who don’t find any effect just put their work away and never submit it for publication- Because it just came back negative- not interesting to most people.
Publication bias:
Research reporting that there is some effect is far more likely to be published than research that shows there is no effect. Not submitting things because they are negative, and not publishing them because they are negative. It’s far more interesting to publish things that have positives.
…and we end up with a much higher probability that a statistically significant effect reported in a published study was just due to random chance.
Regression to Mean
If the first measurement of an effect is extremely high or low compared to the mean, it usually is close to the mean on a subsequent measurement.
And if higher on its second measurement, it will usually have been closer to the mean on the first measurement.
But we might think that there is some causal explanation for the increase of decrease, when in fact, it’s just a natural variation.
Outlyer measurements will tend to go back to the Mean next time we measure them.
File-drawer effect
Researchers who don’t find any effect just put their work away and never submit it for publication- Because it just came back negative- not interesting to most people.
Publication bias
Research reporting that there is some effect is far more likely to be published than research that shows there is no effect. Not submitting things because they are negative, and not publishing them because they are negative. It’s far more interesting to publish things that have positives.
