“…If α(x) ≤ f ′ (x) ≤ β(x) for any x ∈ [a, b] and α, β ∈ C[a, b] then by replacing x = b ∈ [a, b] in (5), the the error of nonstandard quadrature I 6 ( f ) can be bounded as∫ b a (t + a − 2b) β(t) dt ≤ (b − a) ( 2 f (a) − f (b) a − 2b) α(t) dt .(22)Proof. Again, to prove (22) we need to use the results of both theorems 2 and 3 simultaneously such that by replacing x = a in (10) we first obtain(b − a) ( 2 f (a) − f (b) a − 2b) α(t) dt ,(23)provided thatα( t) ≤ f ′ (t) ∀t ∈ [a, b].On the other hand, replacing x = a in(12) gives∫ b a (t + a − 2b) β(t) dt ≤ (b − a) ( 2 f (a) − f (b) ) − ∫ b a f (t) dt ,(24)provided that f ′ (t) ≤ β(t) ∀t ∈ [a, b]. Therefore, combining two latter results (23) and (24) yields (22).…”