Any project dealing with a large corpus, even big data, needs to construct its object of research. This often means defining it according to precise and telling criteria. Quantitative research also means careful selection. Sampling is the key. But how is one to sample within a canonical corpus such as that of the novels of “the Queen of Crime”? There are obviously some numeric indexes: sales figures, number of editions, number of translations. But what else? An interesting ranking of the 78 novels by Agatha Christie can be found at https://agathachristiereader.wordpress.com/christie-index/; Similarly a Goodreads ranking, based on readers votes can be accessed at http://www.goodreads.com/list/show/2126.Best_Agatha_Christie_Book. And there is yet another, more subjective one, below.

Here is its “Top Ten”:

#10. Hercule Poirot’s Christmas (1938)

#9. The ABC Murders (1936)

#8. Cards on the Table (1936)

#7. Death in the Clouds (1935)

#6. Endless Night (1967)

#5. A Murder is Announced (1950)

#4. Crooked House (1949)

#3. Murder on the Orient Express (1934)

#2. Five Little Pigs (1942)

#1. And Then There Were None (1939)

And here is another :

The Murder of Roger Ackroyd	1926
Ten Little Niggers  (And Then There Were None)	1939
The A.B.C. Murders	1936
Murder on the Orient Express	1934
Death on the Nile	1937
Five Little Pigs	1942
The Mysterious Affair at Styles	1920
N or M?	1941
The Body in the Library	1942
Death Comes as the End	1944
The Mystery of the Blue Train	1928
Appointment with Death	1938
The Murder at the Vicarage	1930
One, Two, Buckle My Shoe	1940
Destination Unknown	1954
4.50 from Paddington	1957

What is yours ? What are your top, 5, 10, 20 books by Agatha Christie? Thanks for letting us know. We would be very interested.