Donnie, I'm not seeing anything to indicate what Levi's mother's first name was... Only that her last name at birth was Shehan. So shouldn't she, in fact, be listed as "Unknown" Shehan rather than "Shehan Shehan"?
In the 1820 Census, I find Peter M. Mathis had 1 fwm under 5, 1 fwf under 5 and his spouse (unknown Shehan) in his household. I find a death index for a Easter Mathis who died in 1914 at the age of 90. The index lists Peter Mathis as her Father and "Shehan" as her Mother, and gives her approximate DOB as 1824. She is listed as widowed, but her last name at death is given as Mathis. I would expect it likely another child, possibly even 2 could have been born between 1820 and 1824, but the likelihood of other survivng children between Easter ca. 1824 and Levi ca. 1838 appears to be kinda slim since the 1840 Census only shows 1 male under 5(Levi), 1 female under 5(unknown), and a spouse 20-30(Esther b. ca. 1810?) I have not yet found him in the 1830 Census, that could answer alot, but as of now, would have to assume the female under 5 either did not survive, or could possible have been 4-4 1/2 and married just before the 1850 Census?
1820 - https://familysearch.org/ark:/61903/1:1:XHL7-T68
1840 - https://familysearch.org/ark:/61903/1:1:XHYZ-G53
Death Index - https://familysearch.org/ark:/61903/1:1:F3WX-699
A lot that doesn't make sense to me though... 1) why is daughter Easter Mathis listed as Parker on your profile? 2) what's the liklihood that Peter named a daughter after his 2nd spouse 15 years before he married her? 3) What makes you think Esther of 1850 is not actually Esther Shehan, Peter's first and only wife?