• value: the scale
• score: raw data of individual
• grouped/class intervals: range of values
• greek = population / roman = sample
  • Σ sum what follows
  • N sample size
• average: one value that represents group of scores

frequency distribution listing of values from highest to lowest

• frequency shows how frequently each value occurred
• cumulative: frequency of particular value
• cumulative percentage: cumulative frequency expressed as percentage
• relative: ratio of observed frequency relative to total
• relative percentages: se me fue
• population vs sample es que sacas el grupo de populacion para ser sample
• mean is affected by adding extreme numbers outside overall pattern aka outliers
  • tonto pinche meme
  • median is resistant
  • mean is not resistant
• to find median if number is even, take two middle values and take their average

variability es lugar en donde hay gran tendencia [ ejemplo ]

• measures of central tendency (mean, median, mode)
• variability (range, standard deviation, variance)
• range: length of smallest interval which contains data
  • estimate of how different highest and lowest scores are from one another
  • r = high score - low score
• standard deviation: √ variance
  • variance = standard deviation²
  • [ ejemplo ]

probability helps give educated guesses about population from sample

• p(a) es probabilidad [1] y p(a') es anti-probabilidad [0]
• si p(a) va a pasar es 1
• p(a) + p(a') = 1

normal distribution/curve es cuando most scores are in middle and fewer in extremes

• unimodal (mean = median = mode)
• symmetrical in middle about mean/median/mode
• asymptotic tails never reach x-axis
• completely described by mean and standard deviation
• people fall into +3 and -3 standard deviations from the center
• all area under curve is 1, each half is 0.5 o 50%
• +1 standard dev es 34.13%, +2 es 13.59%, +3 es 2.14%

z-score distance raw score is from mean when measured in standard deviations

• (x - mean) / s

estimation and inference how we draw conclusion from our data

• statistical: predicting unknown population parameters based on sample statistics
• point estimate is mean we used to make inferences
  • unbiased estimator : mean value that is equal to parameter it estimates
  • standard error of mean : standard deviation of sample means in sampling distribution of means
  • standard dev + variance are biased therefore need to N-1, mean is never biased
• confidence intervals: sample mean - 1 standard error to sample mean + 1 standard error

hypothesis es educated guess

• null: statement of equality; no difference
• non-directional research: there is difference, but direction of difference isnt specified
  • class a math =/= class b math
  • no dice cual es mejor o peor
• directional research: there is specified direction of difference
  • class a math > class b math
• significance testing
  • type i error: probability of rejecting null hypo when its actually true
  • type ii error: probability of failing to reject null when its actually false
  • power: probabilty that we correctly reject null

one-sample z-test examines difference between sample mean and population mean

• when testing only one group, observations are independent, variable is normally distributes in population, variance of variable is known
  • state null hypothesis
  • state research (alternative) hypothesis
  • set significance level / probability of type i error
  • state rejection rule

two-sample t-test test if two sample means differ from one another, there are two kinds of this test

• independent samples: differences between two unrelated groups
  • (x1-x2) - (u1-u2) / standard error
  • if sample means are the same something will equal zero
  • degrees of freedom is number of scores in sample that are free to vary
• standard error between means formula
• rejection rule is | t(obs) | > t(crit)

one-way anova es un putero de grupos y un chingazo de trabajo como los otros previos pero en drogas

• quantitive rather than categorical/qualitive data
• research question is about group differences
• 3+ groups
• 'between subjects' anova should be only independent samples
• difference with independent sample is that it has just one null hypothesis, anova has multiple
• asking whether there is difference among means
• within-group variability is within a whole / between-group variability is amongst two
• within each group everyone gets same treatment
• f-ratio of 1 leads to retaining null because differences are due to chance
• three types of deviation scores
  • total
  • between-group
  • within-group
• null hypothesis es 'omnibus'
• f-value always positive, must be greater than 1 to reject null hypo