Quiz 8

Question 1

What does the unique function do in R?

Answer 1

Tells you what all of the unique values are in a particular column
ex. unique(ldt_data$RT_speed)

Question 2

What does the “filter” function do?

Answer 2

Selects the rows that match given criteria
ex. filter(ldt_data,RT_speed==“fast”)

Question 3

What does == do in R?

Answer 3

(two equals signs)
tests/checks whether something is equal
ex. x==y will test if they are equal, say “TRUE” or “FALSE”

Question 4

How do you eliminate rows in a column that have a specific value?

Answer 4

filter(ldt_data, Length_type != “mid”)

Question 5

How can you get R to convert what it is considering characters to be seen as numbers (ex. if your numeric data has quotes around it)

Answer 5

ex. x<- (“10”, 20, 30)—will read all as characters
as.numeric(x) (now can run numeric functions)

Question 6

How can you combine multiple conditions when filtering in R?

Answer 6

use ‘&’ to separate conditions
ex. filter(ldt_data, RT_speed==“slow” & Length ==9)

Question 7

What does the “mutate” function do?

Answer 7

create a new variable (“mutate” data frame)

Question 8

How can you save a new data frame that has the changes made from a mutate operation?

Answer 8

can reassign variable ex. ldt_data
ldt_data<- mutate(ldt_data, Length_10= Length*10)

or, if not appending previous data, can save as new variable:
ex. new_length<-(ldt_data$Length*10)

Question 9

What is the purpose of descriptive statistics?

Answer 9

• to make a summary of the data we have at hand
• to explore existing data for patterns

Question 10

What are the types of data in descriptive statistics?

Answer 10

1. nominal (values are names or labels)
2. ordinal (nominal categories are ordered)
3. interval (equal intervals on scale represent equal differences between the points on the scale)
4. ratio (similar to interval ratios, but here a zero is meaningful, not arbitrary)

Question 11

Describe nominal data

Answer 11

• values simply names or labels
• nominal variables represent the least precise and informative level of measurement in comparison to other data types
  ex. speaker of a language can be:
• “native” or “non-native”
• male or female
  A doesn’t = B (vowel category: i, e, a, o, u)

Question 12

Describe ordinal data

Answer 12

• dealing with ordinal variables when nominal categories are ordered
• ex. “likert scale”-> ‘strongly disagree->’disagree>’agree>’strongly agree’
• ex. vowel height: low—low-mid—high-mid—high

Question 13

Describe interval data

Answer 13

• dealing with interval variables if equal intervals on the scale represent equal differences between the points on the scale
• ex. temperature (C or F)— 25C to 30C same as 20C to 25C (0C doesn’t mean no temperature)
• multiplication, division don’t make sense (20C not twice as warm as 10C)
• negative values possible
  A+B, A-B (birth year 1985, 1978, 2005)

Question 14

Describe ratio data

Answer 14

• ratio variables similar to interval, but zero is meaningful
• zero can mean absence of the thing being measured ex. 0 occurrences of word “aardvark” in a text
• multiplication and division make senes (book can have 2x as many “aardvark” references as another
• negative value not possible
  A x B, A / B (vowel duration: 50, 49, 53, 60ms)

Question 15

What is a pipe in tidyverse?

Answer 15

a special operator %>% to feed things to the next process
ex. df4<-ldt_data %>%
filter(Length>9) %>%
(takes whatever comes before it and passes it on to next step)

Question 16

What kinds of data are considered “categorical”?

Answer 16

nominal and ordinal

Question 17

What kinds of data are called “continuous/numeric” data?

Answer 17

interval and ratio

Question 18

When is it useful to use the mean of a variable?

Answer 18

when the data is normally distributed

Question 19

What is standard deviation?

Answer 19

• measure of how far apart the data points of a variable are from the mean
• indicator of how reliable the mean is
  *when 2 means are the same, the one with the smaller SD is more reliable
  **necessary to include SD when mean is reported

Question 20

What is the median in a list of data points?

Answer 20

the data point that is right in the middle after data points are sorted in ascending order

Question 21

What is the IQR?

Answer 21

the interquartile range
the difference between the 75th percentile (Q3) and the 25th percentile (Q1) of the datapoints (see slide 18 (Descriptive data slides)

Question 22

How do you calculate the IQR?

Answer 22

1. calculate overall median
2. calculate the median of the first half (Q1)
3. calculate the median of the 2nd half (Q3)
4. find difference between Q1 and Q3
   (see slide 18 of descriptive data slides)

Question 23

How are IQR and median related

Answer 23

IQR should be reported with median
smaller IQR= more accurate median

Question 24

When is it better to report mean vs median?

Answer 24

• mean a better choice when the distribution is symmetric and normal
• median better when the distribution of the data is asymmetric and non-normal (skewed) (ex. mean income for a Canadian could seem really high because of small number of very rich people…so data is not normal distribution)—accounts for outliers that are not part of main trend being examined

Question 25

What is a histogram?

Answer 25

a plot that shows the distribution of datapoints in a variable by plotting how many times each datapoint occurs in the data

Question 26

How can you decide whether to use mean or median by looking at a histogram?

Answer 26

A histogram will be bell-shaped if there is a normal distribution, in which case the mean is a better option

Question 27

What are the characteristics of a (perfect) normal distribution?

Answer 27

• mean= median
• 50% of values on the left side of the mean (and median)
• 50% of values on the right side of the mean (and median)
• will be bell-shaped (“bell curve”)

Question 28

What does a left-skewed distribution look like?

Answer 28

• the curve has a long left tail (negatively skewed)