Google Ngram can show graph of usage of a word/phrase based on usage of the word in books or writings with respect to timeline. But is there a way to show such graph with respect to region i.e. showing percentage of people uses the word/phrase in different English speaking areas? Google Ngrams seems to be not supporting such feature.
2 Answers
Try out GloWbE, The Corpus of Global Web-Based English. Its data is divided into twenty English-speaking countries; see here for a list of which countries are included and what their abbreviations are.
For example, I could investigate the distribution of the spellings color and colour. For this sort of comparison, we'll want to go into options and choose PER MIL (per million) rather than RAW FREQ, because the amount of data for each country varies. Here's what I found:
Occurrences per mil in GloWbE of colour
New Zealand 74.35
Australia 72.49
South Africa 70.25
Great Britain 68.22
Ireland 60.93
Singapore 59.83
Malaysia 56.09
Canada 53.00
Jamaica 45.94
Sri Lanka 45.62
Hong Kong 44.75
India 38.26
Tanzania 34.76
Kenya 33.02
Pakistan 32.28
Bangladesh 29.45
Nigeria 28.26
Ghana 26.08
Philippines 15.79
United States 9.49
Occurrences per mil in GloWbE of color
Philippines 93.71
Hong Kong 86.60
United States 80.47
Singapore 77.46
Bangladesh 74.36
Malaysia 64.09
India 62.94
Pakistan 56.11
Jamaica 47.76
Sri Lanka 44.31
Tanzania 41.70
Canada 38.97
Ghana 30.85
Kenya 28.96
Nigeria 27.53
South Africa 26.12
New Zealand 21.94
Australia 17.42
Ireland 15.81
Great Britain 14.97
As we can see, the distribution isn't exactly binary, but there are clear trends, with the U.S. and Great Britain falling roughly on opposite ends of the scale. You can perform similar analyses using GloWbE.
Google Books Ngrams shows the usage in English/British English/American English by selecting from corpus.
