PASS Revision Session

Which type of graph do we use for numerical data?

• Bar chart

• Histogram

• Boxplot

• Clustered bar chart

How do we tell whether we have violated the assumption of constant standard deviation?

• Look at the histogram and see if residuals follow a normal bell curve distribution

• Look at the residual plot and see if there are equivalent data points on the left and right

• Look at the residual plot and see if there are equivalent data points above and below the horizontal axis

• Check to see if the scatterplot shows a positive or negative relationship

It is thought that 18.5% of all smartphone users have an iPhone. To test whether this is true, we sampled 100 smart phone users and found that 23 of them had iPhones. Calculate the test statistic with the relevant formula (to 3 decimal places)

(a)  

Which of the following excel commands will give us the correct p-value for the previous question?

• =2*NORM.DIST (-1.159,0,1,TRUE)

• =2*(1-NORM.DIST(1.159, 0, 1, TRUE))

• =CHISQ.DIST.2T(1.159, 99)

• =1-NORM.DIST(1.159,0,1, TRUE)

I'm interested in seeing how many hours PASS students have spent studying each week for STAT1250. To test this, I randomly select PASS students from the 11am morning session. Is this sample representative, random, both or neither?

• Possibly represenative, not random

• Both

• Neither

• Random but not representative

Calculate the R-squared value for the ANOVA table. Answer as a percentage to 2 decimal places.

(a)  

Number of coffees consumed (in cups) is a _______ variable. Amount of coffee consumed (in mls) is a ________ variable

• 1) nominal

  2) ordinal

• 1) categorical

  2) numeric

• 1) nominal

  2) continuous

• 1) Discrete

  2) continuous

Using the regression output shown, calculate the slope (b1) Answer to 2 decimal places.

(a)  

What does a box plot show that is hard to tell from a histogram? What does a histogram show that is hard to tell from a boxplot?

• Boxplot: median

• boxplot: Variance/spread

• Histogram: Shape of the distribution

• Histogram: outliers

• One sample t-test

  The proportions are not as stated in the null hypothesis

• One sample z-test

  The population mean is not the same as our hypothesised mean

• One sample t-test

  The population mean is not equal to the hypothesised mean

• Paired t-test

  The mean of difference scores is not 0 in the population