📝 Midpoint vs trapezoidal rule accuracy (38 MCQs)

📖 Explanation: For a concave down function, the trapezoidal approximation uses chords that lie below the curve, resulting in an underestimate ($T_n < I$). The midpoint approximation uses tangent lines that lie above the curve, resulting in an overestimate ($I < M_n$). Therefore, the correct ordering is $T_n < I < M_n$. This is a direct Easy of the geometric interpretation of these approximation methods and their error characteristics.

📖 Explanation: For a concave up function, the midpoint approximation lies below the graph ($M_n < I$), so $E_M = I - M_n > 0$. The trapezoidal approximation lies above the graph ($T_n > I$), so $E_T = I - T_n < 0$. This opposite error direction is a key Medium distinction between the two methods. The signs of the errors reflect the geometric relationship between the approximating shapes and the curve.

📖 Explanation: The student's claim is false. While Theorem 7.7.1 states that $|E_M| < |E_T|$ when the function maintains a single concavity on the interval, this guarantee disappears if the function changes concavity. In such cases, the error from the midpoint method could potentially be larger than the trapezoidal error. The relative accuracy depends on the specific function and interval, not just the method itself.

📖 Explanation: Simpson's rule $S_{20}$ is computed as a weighted average of the midpoint and trapezoidal approximations: $S_{20} = \frac{1}{3}(2M_{10} + T_{10}) = \frac{1}{3}(2(0.692835360) + 0.693771403) = \frac{1}{3}(1.38567072 + 0.693771403) = \frac{1}{3}(2.079442123) = 0.6931473743$. This value is much closer to the true value of $\ln 2 \approx 0.693147181$ than either $M_{10}$ or $T_{10}$, demonstrating the superior accuracy of Simpson's rule.

📖 Explanation: A positive second derivative indicates that the function is concave up. For a concave up function, the chord (line segment connecting two points on the curve) lies above the graph. Since the trapezoidal rule sums the areas of trapezoids whose tops are these chords, the approximation overestimates the true integral. The key Medium link is between the sign of the second derivative and the position of the chord relative to the curve.

📖 Explanation: For $f(x) = x^3$, the second derivative is f&#039;&#039;(x) = 6x, which is positive on $(0,2)$. Therefore, the function is concave up on the interval. For a concave up function, the midpoint approximation is an underestimate ($M_n < I$) and the trapezoidal approximation is an overestimate ($T_n > I$). This is a direct Easy of the theorem relating concavity to the direction of the approximation errors.

📖 Explanation: For a concave down function, the midpoint approximation is an overestimate ($I < M_n$) and the trapezoidal approximation is an underestimate ($T_n < I$). Given $M_{10} = 2.5$ and $T_{10} = 2.7$, we have $T_n < I < M_n$, which translates to $2.7 < I < 2.5$. This implies $2.5 < I < 2.7$, meaning the exact value must lie between the two approximations. This is a direct Easy of the error bounds and their geometric interpretation.

📖 Explanation: The error bound for the midpoint approximation is $|E_M| \le \frac{(b-a)^3 K_2}{24n^2}$, while for the trapezoidal approximation it is $|E_T| \le \frac{(b-a)^3 K_2}{12n^2}$. The key observation is that the trapezoidal error bound is exactly twice the midpoint error bound, assuming the same $K_2$. This is why the midpoint approximation is generally more accurate than the trapezoidal approximation for functions with consistent concavity.

📖 Explanation: The error is defined as $E = I - \text{approximation}$. Thus, $E_M = I - M_5 = 0.841470985 - 0.842875074 = -0.001404089$. Similarly, $E_T = I - T_5 = 0.841470985 - 0.838664210 = 0.002806775$. The negative sign for $E_M$ indicates that the midpoint approximation is an overestimate, which is consistent with the function $\cos x$ being concave down on $(0,1)$.

📖 Explanation: For a concave down function, $M_n$ is an overestimate and $T_n$ is an underestimate. Therefore, we expect $M_n > I > T_n$. The only option that satisfies this ordering is $M_{10} = 4.2 > I = 4 > T_{10} = 3.9$. The other options either have the wrong order or both values are on the same side of the exact value, which would violate the known error direction for a concave down function.

📖 Explanation: If f&#039;&#039;(x) = 0 on $[a,b]$, then the function is linear. Both the midpoint and trapezoidal approximations are exact for linear functions. The trapezoidal rule uses chords that coincide with the linear function, and the midpoint rule's tangent line also coincides with the function. Therefore, both approximations equal the exact integral for any $n$. This is a special case where the concavity-based error guarantees become exact equalities.

📖 Explanation: For a positive second derivative (concave up), the midpoint approximation is an underestimate ($M_n < I$), so $E_M = I - M_n > 0$. The trapezoidal approximation is an overestimate ($T_n > I$), so $E_T = I - T_n < 0$. Thus, the errors have opposite signs, with $E_M$ being positive and $E_T$ being negative. This relationship is a direct consequence of the geometric interpretation of the two approximation methods.

📖 Explanation: This is a special case. If f&#039;&#039;(x) = c (a constant), the function is a quadratic polynomial. For quadratic functions, the relationship $|E_T| = 2|E_M|$ holds exactly for any $n$. This is because the error in Simpson's rule is zero for quadratics, and Simpson's rule is a weighted average of the midpoint and trapezoidal rules. In general, this exact relationship does not hold; it's only an error bound relationship. The statement is true only for functions with a constant second derivative.

📖 Explanation: While option C is true for concave up functions, it is not a general reason for the midpoint method's superior accuracy. The primary reasons are: (1) tangent lines approximate the curve better than chords when concavity is consistent, (2) the error bound for the midpoint method is smaller, and (3) within each subinterval, the midpoint approximation's error often cancels out to some degree because the function is overestimated on one side of the midpoint and underestimated on the other. Option C only describes the behavior for a specific concavity type and doesn't explain the general accuracy advantage.

📖 Explanation: For $f(x) = x^2$, f&#039;&#039;(x) = 2 > 0, so the function is concave up. Therefore, $|E_M| < |E_T|$ for any $n$. Specifically, $M_2$ will be an underestimate and $T_2$ will be an overestimate, but the error in $M_2$ will be smaller. This can be verified by computing the exact value $\int_0^2 x^2 dx = \frac{8}{3} \approx 2.6667$, and the approximations. The student should recognize the concavity and apply the theorem about the relative accuracy of the two methods.

📖 Explanation: For a concave up function, the midpoint approximation is always an underestimate and the trapezoidal approximation is always an overestimate, regardless of the number of subintervals. Therefore, the inequality $M_n < I < T_n$ holds for any $n$. Thus, if $M_{10} < T_{10}$, then $M_{20} < T_{20}$ will also hold. The approximations will both get closer to $I$ as $n$ increases, but the ordering remains the same. This is a fundamental property of the methods for functions with consistent concavity.

📖 Explanation: Cubic polynomials have a non-zero third derivative, so neither the midpoint nor the trapezoidal approximation is exact in general. The midpoint and trapezoidal rules are exact only for linear functions (and the midpoint rule is also exact for some quadratics under certain conditions, but not generally for cubics). Simpson's rule, however, is exact for cubic polynomials because it uses parabolic arcs and the error term involves the fourth derivative.

📖 Explanation: For a function with a positive second derivative (concave up), the trapezoidal approximation $T_n$ overestimates the exact integral $I$. Therefore, the error $E_T = I - T_n$ is negative. This is a direct Easy of the geometric interpretation of the trapezoidal rule: the chords lie above the curve, leading to an overestimate. The sign of the error is determined by the concavity of the function.

📖 Explanation: The inequality $|E_M| < |E_T|$ indicates that the absolute error of the midpoint approximation is less than the absolute error of the trapezoidal approximation. This means the midpoint approximation is more accurate (closer to the true value) than the trapezoidal approximation. This is a key result for functions with consistent concavity, and it forms the basis for the weighted average used in Simpson's rule.

📖 Explanation: A positive second derivative indicates the function is concave up. For a concave up function, the tangent line at the midpoint lies below the curve. Since the midpoint approximation uses rectangles whose heights are the function values at the midpoints, this is equivalent to using tangent lines, resulting in an underestimate. This geometric interpretation is crucial for understanding the direction of the error.

📖 Explanation: The theorem stating $|E_M| < |E_T|$ assumes the function maintains a single concavity on the interval. If the function changes concavity, this guarantee no longer holds. In such cases, the errors from the two methods can behave differently, and it is possible for the trapezoidal approximation to be more accurate than the midpoint approximation for a particular function and interval. This highlights the importance of the concavity condition.

📖 Explanation: The student's claim is false because the trapezoidal approximation is an overestimate for concave up functions and an underestimate for concave down functions. The best counterexample is a concave up function, where the trapezoidal rule overestimates. $f(x) = -x^2$ is concave down, so the trapezoidal rule is an underestimate, which would support the student's (incorrect) claim. The correct counterexamples are $f(x) = x^2$ or $f(x) = e^x$, which are concave up and thus the trapezoidal rule is an overestimate.

📖 Explanation: The error bounds for both the midpoint and trapezoidal approximations depend on the maximum value of the second derivative, denoted as $K_2$, on the interval of integration. Specifically, $|E_M| \le \frac{(b-a)^3 K_2}{24n^2}$ and $|E_T| \le \frac{(b-a)^3 K_2}{12n^2}$. The second derivative measures the curvature of the function, which directly affects the accuracy of these linear approximations.

📖 Explanation: For a concave down function, the trapezoidal approximation is an underestimate ($T_n < I$) and the midpoint approximation is an overestimate ($I < M_n$). Additionally, for functions with consistent concavity, the absolute error of the midpoint approximation is less than that of the trapezoidal approximation ($|E_M| < |E_T|$). Thus, the correct combination is $T_n < I < M_n$ and $|E_M| < |E_T|$. This question tests the understanding of both the direction of the errors and their relative magnitudes.

📖 Explanation: The function $f(x) = \sqrt{1+x^2}$ has a positive second derivative (it is concave up). For a concave up function, the midpoint approximation is an underestimate ($M_n < I$) and the trapezoidal approximation is an overestimate ($T_n > I$). Therefore, the ordering is $M_n < I < T_n$, or equivalently $T_n > I > M_n$. This is a direct Easy of the concavity-based error direction theorem.

📖 Explanation: The midpoint rule's superior accuracy stems from the cancellation of errors within each subinterval. The tangent line at the midpoint overestimates the function on one half of the subinterval and underestimates it on the other half. These errors partially cancel, leading to a smaller net error. In contrast, the trapezoidal rule's chord lies entirely above or below the curve (depending on concavity), so the error does not cancel in the same way. This is a key insight into why the midpoint rule often outperforms the trapezoidal rule.

📖 Explanation: A negative second derivative indicates the function is concave down. For a concave down function, the chord connecting two points on the curve lies below the graph. Since the trapezoidal rule uses these chords, it underestimates the true integral. This is the geometric interpretation of the error direction for a concave down function. The student must connect the sign of the second derivative to the concavity and then to the position of the chord relative to the curve.

📖 Explanation: For a concave down function, the midpoint approximation is an overestimate ($M_n > I$) and the trapezoidal approximation is an underestimate ($T_n < I$). Therefore, we need $M_{10} > 10 > T_{10}$. The only option that satisfies this is $M_{10} = 10.2$ and $T_{10} = 9.8$. This tests the student's ability to apply the concavity-based error direction to a concrete numerical example.

📖 Explanation: The error bound for the trapezoidal approximation is $|E_T| \le \frac{(b-a)^3 K_2}{12n^2}$, while for the midpoint approximation it is $|E_M| \le \frac{(b-a)^3 K_2}{24n^2}$. Comparing the denominators, the trapezoidal error bound is exactly twice the midpoint error bound: $\frac{1}{12} = 2 \times \frac{1}{24}$. This means the trapezoidal method has a theoretical error bound that is twice as large, suggesting it is generally less accurate.

📖 Explanation: For a concave up function, the tangent line at the midpoint lies below the curve, so the midpoint approximation is an underestimate ($M_n < I$). The chord used in the trapezoidal rule lies above the curve, so the trapezoidal approximation is an overestimate ($T_n > I$). Combining these gives the ordering $M_n < I < T_n$. This is a direct Easy of the geometric interpretation of the two methods based on the function's concavity.

📖 Explanation: For a concave down function, the midpoint approximation is an overestimate ($I < M_n$) and the trapezoidal approximation is an underestimate ($T_n < I$). Therefore, the exact integral must lie between the two approximations: $T_n < I < M_n$. Given $M_{10} = 3.1$ and $T_{10} = 3.3$, we have $3.3 < I < 3.1$, which is a contradiction. This indicates the student's data is inconsistent with a concave down function. The correct relationship for concave down is $T_n < I < M_n$, so if $M_{10} = 3.1$ and $T_{10} = 3.3$, the function cannot be concave down. The student should recognize that the order of the approximations is reversed for a concave down function.

📖 Explanation: For a positive second derivative (concave up), the midpoint approximation is an underestimate ($M_n < I$), so $E_M = I - M_n > 0$. The trapezoidal approximation is an overestimate ($T_n > I$), so $E_T = I - T_n < 0$. Therefore, $E_M$ is positive and $E_T$ is negative. This is the correct sign relationship for a concave up function. The student must link the concavity to the direction of the errors for both methods.

📖 Explanation: Both the midpoint and trapezoidal approximations are exact for linear functions. For $f(x) = x$, the second derivative is zero, so the function is linear. The chords used in the trapezoidal rule coincide with the function, and the tangent lines used in the midpoint rule also coincide with the function. Therefore, the error is zero for any $n$. Quadratic and cubic functions will have non-zero errors in general, and trigonometric functions are not linear.

📖 Explanation: For a concave up function, $E_M > 0$ and $E_T < 0$. Additionally, from the error bounds, $|E_T| \le 2|E_M|$ in the worst case, and generally $|E_T| > |E_M|$. So, $E_T$ is negative and its magnitude is greater than the magnitude of $E_M$ (which is 0.01). Therefore, $E_T < -0.01$. The exact value of $E_T$ cannot be determined without more information, but its sign and relative magnitude are known. This tests the understanding of both the sign and the relative magnitude of the errors.

📖 Explanation: The student's argument is flawed because using more points does not guarantee higher accuracy. The key counterargument is that the error bound for the trapezoidal approximation is twice as large as that for the midpoint approximation (for functions with consistent concavity). The midpoint approximation's use of tangent lines at the midpoints often provides a better fit to the curve than the chords used in the trapezoidal rule, leading to smaller errors. The number of function evaluations is not the primary factor in the relative accuracy; the shape of the approximating function is.

📖 Explanation: For functions with a constant second derivative (quadratic functions), the relationship $E_T \approx -2E_M$ holds exactly. This is the basis for Simpson's rule, which combines the two approximations to cancel the error. For general functions, this relationship is only approximate, and it becomes more accurate as $n$ increases. The relationship is not true for all functions, but it is a useful approximation that motivates the weighted average used in Simpson's rule.

📖 Explanation: For a concave up function, the midpoint approximation is an underestimate ($M_n < I$) and the trapezoidal approximation is an overestimate ($T_n > I$). Given $M_{10} = 1.5$ and $T_{10} = 1.7$, we have $1.5 < I < 1.7$. This means the exact integral is between the two approximations. The student should not simply average the two values; the exact value could be anywhere in the interval (1.5, 1.7), but it is guaranteed to be between them. This tests the Easy of the concavity-based ordering to determine the range of the exact integral.

📖 Explanation: The error in the midpoint approximation is defined as the exact value minus the approximation: $E_M = I - M_n$. This definition allows for both positive and negative errors, indicating whether the approximation is an overestimate or underestimate. The absolute error is $|E_M| = |I - M_n|$. The definition $E_M = M_n - I$ is sometimes used, but the convention in the text and many calculus textbooks is $E_M = I - M_n$. The student must recall the correct definition of the error.