Psychological Statistics II LAB

Psychological Statistics ■ Q2 - 17.5

● Inferential Statistics ■ Q1 - 16
● Descriptive Statistics - help psychologists ■ Q3 - 19
summarize or describe observations. They ■ IQR = 3
provide this organizing and summarizing function. ● Variance - (Standard deviation)2 The best
○ 2 main types: measure of diversion because it considers all
1. Measures of Central 2 Σ(𝑥−𝑥)
2
scores in the data set. 𝑠 =
Tendency (the typical score) 𝑛−1
2. Measure of Dispersion (the 1. Get the mean
spread of the data) 2. Deduct each score from the mean
3. Get the squared difference
Measures of Central Tendency 4. Sum the squared difference
● Most common summary description of data, 5. Divide the/ sum by n-1
which, as the term implies, indicates the center ● Our unit of measurement is seconds, but there’s
of the distribution scores. no seconds squared. What we do with variance
● Mean, median, and mode. is get the standard deviation.
● Mean is the most powerful and advisable to use, ● Standard Deviation - Square root of the
because it considers the entire data. 2
Σ(𝑥−𝑥)
● The mean is sensitive to outliers. variance. 𝑠 = 𝑛−1
● We cannot use the mean when the data is
skewed because it is not showing an accurate
Analyzing Data
representation of the data.
● The normal distribution is a probability function
● Mode is categorical, or nominal. Used to see the
that describes how the values of a variable are
frequency of certain scores.
distributed. It’s a symmetric distribution where
● Which measure of central tendency should you use?
most of the observations cluster around the
○ The important point to keep in mind when
central peak.
choosing a measure of central tendency is
● 1 parameter, or criteria, we need to abide by is
that it should give you a good indication of
that the data should be normally distributed. This
the typical score in your sample.
means it's a good representation of the data set.
○ After you have calculated your measure
● In statistics, there’s parametric tests and
of central tendency compare it to the
non-parametric tests.
scores in your sample and satisfy
● Parametric tests are more powerful for the null
yourself that it looks representative of
hypothesis, because it’s normally distributed. For
the scores as a whole.
this test, it’s important for the data to be normally
distributed.
Measure of Dispersion
● Measures of dispersion indicate how the scores
Characteristics
are spread out about the center.
● It should be symmetrical about the mean.
● The peak of the bell curve is dependent on the
● The tail should meet the x-axis at infinity.
mean. The wideness of the curve is dependent
● It should be bell-shaped.
on the measure of dispersion.
● Data is deviated for normality for symmetry and
● If the data is dispersed, the curve will be wide,
pointiness.
which is not good because it means the data is
very dispersed from the average.
Skewness
● The outliers show that it’s not a normal
● Positive - the frequent scores are clustered at
distribution.
the lower end and the tail points towards the
● We want a small variability.
higher or more positive scores.
● The range is also susceptible to outliers.
● Negative - the frequent scores are clustered at
● Interquartile Range - Q3 minus Q1. Q3 is the
the higher end and the tail points towards the
upper quartile, and Q1 is the lower quartile. It
lower or more negative scores.
doesn’t consider the entire data.
● Near to 0, the better.
○ Ex. 15, 16, 17, 18, 19, 20
1
 Psychological Statistics II LAB
𝑠
● Equation: 𝑆𝐸 =
Kurtosis 𝑥 𝑛
● Refers to the degree to which scores cluster at the
end of the distribution (the tails) and this tends to Confidence Intervals
express itself in how pointy a distribution is. ● Confidence intervals of the mean are interval
● Positive = Leptokurtic = heavy-tailed distribution. estimates of where the population mean may lie.
It has many scores in the tails (a so-called ● They provide us with a range of scores (an
heavy-tailed distribution) and is pointy. interval) within which we can be confident that
(Leptokurtic Distribution) the population mean lies.
● Negative = Platykurtic = light-tailed distribution.
It’s relatively thin in the tails (has light tails) and The Standard Normal Distribution
tends to be flatter than normal. ● The z-score is expressed in standard deviation
● Near to 0, the better. units: that is, the z-score tells us how many
standard deviations above or below the mean
Histogram our score is.
● A frequency distribution, or histogram, which is a 𝑋−𝑋
● 𝑧=
graph plotting values of observations on the 𝑆
horizontal axis, with a bar showing how many ● Another useful feature is that we can use it to
times each value occurred in the data set. calculate the proportion of the population who
would score above or below your score.

Box and Whisker Plots

● Enable us to easily identify extreme scores as
well as well as seeing how the scores in a
sample are distributed.

Standard Error
● A statistical term that measures the accuracy
with which a sample represents a population.
● In statistics, a sample mean deviates from the
actual mean of a population; this deviation is the
standard error.