Chapter 1 - Basic Concepts Flashcards
what are statistics?
study of collecting, organizing and analyzing data (sample based on a population) to understand patterns and make decisions.
why are statistics important?
to get answers.
eg. what forms of therapy are more effective?
education wise: use statistics in bachelors, masters, PhD, …
do we all have an intuitive grasp of statistics? give an example.
yes we do. say flipping a coin 10 times. getting head 10 on 10 times is far more surprising then 5 out of 10 times.
based of the example you just used, it landed 10/10 times heads. would you say the coin is fair?
no, we would be biased in some way. however, unusual does not mean impossible.
what is the goal of statistics?
estimation
in terms of the jelly bean example, what would be the sample? population?
the sample would be the portion that you decide to grab a handful of. the population is the entirety of the jelly beans in the jar.
what would be more helpful? a small sample or a large sample?
- You find that 28% of the jelly beans are red in your large sample
- Your large sample has 25 jelly
beans and 7 of them are red - 7/25 = .28 = 28%
- How many red jelly beans do you think are in the jar?
- With only our sample, our best
estimate is 28% - Is this would be a good estimate?
- Another sample would likely
have a slightly different
number of red jelly beans - It could have 28%, but it could
also have more or less than 28%
-it really depends just how big of a sample you pick up. if you pick up 95% of the jelly beans, then yes it is more likely to be accurate)
psychology involves measuring things, what do psychologists measure? (3)
-construct: the “thing” to be measured (e.g. health)
-variable: physical or abstract attributes that we wish to measure, can have a specific measure. (e.g. # heart attacks in mtl over the past year)
-score: value that an individual has on a particular variable (e.g. yes or no to having a heart attack)
what is the difference between data and datum?
data is plural
datum is singular
what is an independent variable?
variable that you can manipulate or categorize (e.g. study time - minutes, hours or days)
what is a dependent variable?
variable that you measure (e.g. grade - exam score, term GPA, cumulative GPA, etc.)
what is a confounding variable?
varies systematically with the independent variable making it impossible to determine which drives the effect (e.g. cheating, copying, chat GPT) - not manipulating it but it can influence the independent variable.
what is the difference between qualitative and quantitative?
qualitative = not numerical, things with names and values (e.g., hair colour, flavour of ice cream, etc.)
quantitative = numerical values, numbers
what are the two sub-categories for qualitative variables?
nominal and ordinal
what is a nominal variable?
-names with no natural ordering (e.g. blue is not bigger then orange)
* Sometimes called Categorical
Variables, as values can be
discrete categories
* E.g., Political party, Eye Colour,
Smartphone, Cookie Preference
what is an ordinal value?
(*first version)
- rank order!
- between qualitative and quantitative, they have properties of both
- E.g., Marathon Runners; there
are no equal intervals
between those who finish 1st,
2nd, 3rd, etc - There is a natural order of
scores, but no units of
measurement
what is an ordinal value?
(*second version)
letter grades. natural ordering but no unit! BUT can be converted to numbers and treated as intervals. qualitative!
what are the two sub-categories for quantitative variables?
interval and ratio
what is an interval variable?
-interval scales have equal units of measurement but no zero point.
-ALL ratio scales are equal interval scales, but NOT all equal interval scales are ratio scales.
true or false? not all scales have absolute zeros.
true. kelvin vs celsius vs Fahrenheit
true or false? Fahrenheit is an interval scale, not a ratio scale.
true
what is a ratio variable?
-Ratio Scales have an
absolute zero and a unit of measurement
that corresponds to a constant physical quantity
* Ratio Scales can measure two values in
terms of their relative distance from zero
* E.g., 3ft tall versus 6 ft tall (2Xs from 0)
* Values can be Discrete or Continuous
* Integers/Whole Numbers VS Real Numbers
what is the difference between a integer/whole number and a real number/continuous
integer = whole number (no decimals)
continuous = decimals
recap!! name all four levels of measurements!
nominal, ordinal, interval and ratio.
many variables are not physical (e.g. we can’t measure intelligence with devices like a ruler). so, define psychological constructs and operational measures.
- Psychological Constructs: distinct abstract quantities that explain an aspect of behaviour but cannot be measured directly (e.g. depression)
- Operational Measures: tool used to measure a psychological
constructs.
what is the difference between validity and reliability.
validity: a device is valid if it measures what it is suppose to measure.
reliability: a device is reliable if it produces similar if not identical measurements each time it is applied to the same object.
what is the measurement error?
each time something is measured a slightly different score will be obtained. (nothing can measure without error)
difference between populations and samples.
populations: compromises all scores of a variable (aka parameter)
samples: subset of a population (aka statistics)
define parameters.
parameters are characteristics of the population (real entities)
how are parameters denoted?
greek letters (miew, variants, sigma and rho)
define statistics.
numerical characteristics of a sample (guesses)
how are statistics denoted?
by lower case Latin letters: m, s^2, s, r
what is a sampling bias?
- not all members of
the population had an equal
chance of being selected in the
sample.
-Leads to statistics being poor
estimators of parameters`
what helps reduce sampling bias?
simple random sampling: In a truly random sample each
unit has an equal chance of being sampled
Do we think the scores of the sample will equal the scores of the population?
no! it is highly unlikely. A sample statistic is often never exactly equal to the population parameter.
what is the difference between the population parameter and the sample statistic?
sampling error
observed score = ?
observed score = true score + error
does sampling error = sampling bias?
no! sampling error can not be avoided, but bias theoretically can.
true or false?
1. the smaller the sample, the bigger the sampling error.
2. the bigger the sample, the smaller the sampling error
- true!
- true!