Data Analysis Flashcards
Density (d) formula
d=mass/volume (g/mL)
Specific Gravity
Weight of substance: Weight of equal volume of standard (e.g. water)
Specific Gravity formula
sg=weight substance (g)/ weight equal vol standard (g)
Descriptive statistics
Data in numerical/ graphical form
Inferential statistics
Draws conclusions about pops.
Quantitative variable
Numerical values that can be averaged (e.g. height, weight)
Categorical variable
Categories/ groups that=classification (e.g. eye colour)
QV: Continuous
Can be measured- infinite no. values w/in range (e.g. weight)
QV: Discrete
Can be categorised into classification- based on whole numbers, i.e. only finite range no.s (e.g. no. deaths)
CV: Nominal
Categories do not have ordering (e.g. sexes)
CV: Ordinal
Categories have logical ordering (e.g. severity disease, year levels)
Observational study
Processed observed- data recorded (e.g. blood pressure)
Randomised experimental study
Specific procedure whereby action=controlled + data measured (e.g. drug vs placebo)
Placebo
Treatment that looks the same, but has no therapeutic effect
Case control study
Study where cases w/ particular attribute (e.g. heart disease) is compared to controls who don’t
Addition rule
Prob. event A or B
Multiplication rule
Prob. event A and B
Conditional prob.
P(B|A)- Prob B given A has occurred
Mutually exclusive events
2 events that do not overlap (no common events)- prob. A and B
Independent events
P(B|A)= P(B), knowing that A has already occurred (P(B)= unchanged)
Splitting prob.
Law total prob.- allows prob. event B to be calculated (B=(A and B) or (Ac and B)
Baye’s rule
Allows ‘turn around’ conditional prob. to occur
Sensitivity
Ability of test to correctly identify people who have given disease/ disorder (the more sensitive- the fewer false -ves)
Specificity
Ability of test to correctly exclude individuals who do not have given disease/disorder (the more specific- the fewer false +ves)
Random variable
Assigns no. to each outcome random circumstance/ each unit in pop.
Continuous random variable
Can take any value in interval/ collection intervals
Discrete random variable DRV
Can take countable list of distinct values (integers)
Prob. distribution function (Pdf)
Formula/ table that assigns probs. to all possible values X
Cumulative distribution function (Cdf)
Formula/ table that provides cumulative probs. P(X
Expectations for RVs
Expected value RV= mean value variable X in same sample space possible outcomes
Standard deviation for DRV
Similar to average distance from random variable to its mean
Binomial RV
No. successes (x) in n repeated trials of binomial experiment
Continuous random variable CRV
Outcome can be any value in interval/ collection intervals
Normal random variable NRV
Most common type CRV
Prob. density function
Indicates how densely prob. of conc. about each value
Confidence Intervals (CI)
Use sample data to provide interval values that is believed to cover the true/ unknown value pop. parameter
Sampling distribution
Distribution possible values statistics for repeated random samples same size from pop.
Confidence level
Prob. that procedure used to determine interval will provide interval that includes pop. parameter (=true value)
t-distribution
t-density is similar in shape to standard normal density (symmetric around 0 + bell-shaped)
