In: Statistics and Probability
U2 is a famous band with studio albums that span the 1980’s, 1990’s, and 2000’s. Suppose you have a friend who has not heard of this band but wants to become familiar with their music. Here is a summary of their studio album recordings:
Year Released Album Name Number of Songs
1980 Boy 11
1981 October 11
1983 War 10
1984 Unforgettable Fire 10
1987 The Joshua Tree 11
1988 Rattle and Hum 17
1991 Achtung Baby 12
1993 Zooropa 10
1997 Pop 12
2000 All That You Can't Leave Behind 11
2004 How to Dismantle an Atomic Bomb 11
2009 No Line on the Horizon 11
2014 Songs of Innocence 11
TOTAL 148
Suppose your friend does not have time to listen to every track on their albums, and instead wants to take a sample of their songs to listen to in order to familiarize himself with the music.
Your friend decides he has time to listen to about 40 songs (give or take a few as needed) and wants those songs to be representative of the group’s music.
a. Suppose your friend created an online poll and asked people to name their favorite U2 song, and he listened to the 40 songs that got the most votes. What do you think of this approach in this situation? Explain.
This would be considered as biased sample as human bias is observed in this approach. Those who vote in this poll would vote basis their choices and a certain song or album may be exaggerated in terms of good music or bad music. Hence this would not be a better approach to become familiar with their music.
a. Explain briefly how your friend could choose a simple random sample of 40 songs from U2.
All the songs could be collected in one library and assign an identity to each song. Post that a simple random sampling tool can be used to choose a song. In excel all the song identity could be placed in excel column and than rand function can be use to choose a sample
a. Suppose your friend is concerned with the possibility of the simple random sample choosing too many songs from one era of the band (and therefore not giving a good representation of their music overall). What sampling approach would you recommend for him now? Explain briefly.
We would suggest stratified sampling method to him as it will first categories all the songs in different ERAs which will be considered in stratas and then simple random sample could be chosen from those stratas.
a. Suppose your friend believes strongly that you have to listen to an entire album to really get the sense of the band’s music, because “each album is like its own story”. What sampling method would you suggest to your friend now? Explain a few specifics of how your friend could conduct this approach.
I would suggest the fried to use convenient sampling in this case as it would be convenient to choose any sample considering the time contraints in mind
a. Which approach in parts (a) through (d) would you recommend overall? Explain your choice.
We would suggest stratified sampling method to him as all the samples would be representative of all strata groups and randomness in sample would give required variance to the sample and would be true representative of population