Not sure where you're getting that Martha married at age 8. In the 1910 census (where she was recorded as being 20), it indicates that she'd been married for eight years, calculating back to being age 12 at her marriage.
The marriage record ((https://familysearch.org/ark:/61903/1:1:XRSC-GXN) indicates she was 15 at the time of the marriage -- 6 Jan 1904. This calculates back to an 1889 birth.
And her age in 1910 suggests a birth year of 1890.
(Her gravestone claims she was born June 16, 1885.)(*)
Regarding Cherokee heritage. She very well may have been Cherokee. But she may not have been. One of the first things I experienced in doing family history research was the debunking of a long-held family tradition/understanding. Not that I set out to do that-- I didn't. In fact, I was searching for records that would confirm our family's long held belief that our immigrant couple married on the ship on the way over from Europe in the mid 1800s. What I found (church record of their marriage in Buffalo, NY) debunked it. You should seek records that confirm the race of your ancestress. (Others can speak to how the use of DNA might also help.)
The census records list the race of each person. In the 1910 census, Martha was marked as white.
We're not sure the Martha in the Coley family in 1900 is her. The birth info is off too much for me to be comfortable with it: May 1882. But if the census taker's notes were accurate, providing correct birth information may not have been the family's strong suit. In any case, all members of that family are enumerated as white race.
Ditto in 1920 census : white.(Source: Speedwell, Wythe, Virginia; Roll: T625_1919; Page: 2A;Enumeration District: 191; Image: 260)
Ditto in 1930 census: white
(Source: Wytheville, Wythe, Virginia; Roll: 2465; Page: 1B;Enumeration District: 0013; Image: 915.0;FHL microfilm: 2342199)
And again in 1940 census: white.
(Source: Wytheville, Wythe, Virginia; Roll: T627_4303; Page: 6B;Enumeration District: 99-22)
But if I were just to go with these four different census records (1910, 1920, 1930 and 1940), I see no evidence supporting the Cherokee heritage.
Now, census information is not always accurate. We've clearly learned that. And given the social stigma about race, people could have lied-- about their own race or the race of the people they were enumerating (in small towns, the enumerators probably knew the families).
In any case, if I were you, I'd be wanting to find other records that supported the story of her Cherokee history.
Also the story that she was purchased to be Albert's wife. She was born long after slavery was abolished. The selling of human beings was therefore illegal. I don't know enough about Native American history to know if the abolishment of slavery applied to them as well. I'm assuming it did, but who knows? And may even if it was illegal, the family engaged in it.
In any case, you've got a really interesting history here that should be very fun to continue to explore. If you're into that. (Obviously, I am... ;-)
(*)One more thing: there were TWO different Martha J's who married Grubbs in Wythe Co., VA. See the 1930 census:
- "Yours": District 113, Wythe, Wythe County: married to Albert W. Grubb (This census also indicates she was 15 at marriage.) Source: Wytheville, Wythe, Virginia; Roll: 2465; Page: 1B;Enumeration District: 0013; Image: 915.0;FHL microfilm: 2342199
- Another Martha J: District 4, Black Lick, Wythe Co.: married to William E. Grubb; she's 46 calculating back to an 1884 birth year; she was 19 at marriage calculating to a 1903 marriage year.