Tabulating and Plotting Data Flashcards
what are two ways to summarize our numbers?
- frequency distribution (table that shows classes-intervals of data with a count of the number of entries in each interval
- graph (preferable because of pictorial representation)
frequency
the number of data entries in that interval
lower class limit
smaller number in interval
how to find class width
if table given, subtract vertically
if only data given, (max#-min#/# of classes) and round to nearest integer
how to find class limits from data given
lowest number=lowest limit, add class width to find lower limit of the next class/interval. Upper limit is just one less than the lower limit of the next class/interval.
frequency histogram
bar graph that represents the frequency distribution
x-axis=
quantitative data values
y-axis=
frequency
requirement of frequency histogram:
bars must be touching
class boundaries
the numbers that separate classes on a frequency histogram, must be consecutive to allow for no gaps and touching bars)
to find class boundaries:
+0.5 to upper limites and -0.5 for lower limits
how to find midpoint of a class
(lower class limit + upper class limit)/2, this can be used to plot frequencies which changes label on x-axis but not the distribution
positive skew
“trending” / stretching toward positive x-axis
negative skew
“trending” / stretching toward negative x-axis
frequency polygon
a line graph that emphasizes continuous changes in frequencies on midpoint lables
how to construct frequency polygon from frequency distribution table?
line must start/end on x-axis, extend both ends by one class width past the first class midpoint
what is a pro of using stem-and-leaf plots
still able to see original data values
unimodal symmetrical
one mode or “peak,” has a centre of symmetry
bimodal symmetrical
two modes or “peaks” of roughly equivalent height, has a centre of symmetry
what is the number of peaks/distribution type based on?
the number of peaks observed on the graph, based on what it looks like. if the peaks are super close, in the same class or interval, it may look like one peak and thus be unimodal.
negative skew is synonymous with:
left skew
positive skew is synonymous with:
right skew
kurtosis
extent of deviation from normal curve in width of curve and thickness of tails
mesokurtic curve
“middle-ground” - this is the normal distribution
leptokurtic curve
“tall and skinny” - tall and skinny peak, shortened tails
platykurtic curve
“flat and wide” - lower peak and lengthened tails
what is the goal of graphs?
to provide an accurate visual representation of your results
what constitutes as cheating graphs that distort data?
truncated graphs without line break, no error bars (hard to confirm statistical significance), failing to adjust for distorting confounds (ex. correcting for population differences)