Astronomy Quiz: Measurement Uncertainty

20 questions · exam conditions

0:00

Measurement UncertaintyQuestion 1 of 20

In the analysis of a stellar spectrum, the abundance of iron is determined to be $[Fe/H] = -0.50 \pm 0.05$ from statistical sources of error. The astronomer also identifies a potential systematic error from the atmospheric model used, which could shift the result down by 0.10. How should the astronomer report the final result and its uncertainty?

Combine the errors to report

[Fe/H] = -0.50 \pm 0.15

Adjust the central value and report

[Fe/H] = -0.60 \pm 0.05

Average the two errors and report

[Fe/H] = -0.50 \pm 0.075

Report the result as

[Fe/H] = -0.50 \pm 0.05_{stat} \pm 0.10_{sys}

← Back to quizzes

Astronomy Quiz

Astronomy Quiz: Measurement Uncertainty

Practice Measurement Uncertainty in Astronomy with focused quiz questions that help you check what you know, review explanations, and build confidence with test-style prompts.

What this quiz covers

This quiz focuses on Measurement Uncertainty, giving you a quick way to practice the rules, question types, and explanations that matter most for Astronomy.

How to use this quiz

Try each quiz question before looking at the correct answer. Use the explanations to review missed ideas, then come back to similar questions until the pattern feels familiar.

All questions

Question 1

Combine the errors to report $[Fe/H] = -0.50 \pm 0.15$ .
Adjust the central value and report $[Fe/H] = -0.60 \pm 0.05$ .
Average the two errors and report $[Fe/H] = -0.50 \pm 0.075$ .
Report the result as $[Fe/H] = -0.50 \pm 0.05_{stat} \pm 0.10_{sys}$ . (correct answer)

Explanation: When analyzing stellar spectra, you'll encounter two distinct types of measurement uncertainty that must be handled differently. Statistical errors arise from random variations in your data (like photon noise), while systematic errors come from consistent biases in your measurement method (like flawed atmospheric models or calibration issues). The key principle is that these error types should never be combined algebraically because they have different sources and characteristics. Statistical errors follow well-understood probability distributions and can be reduced by collecting more data. Systematic errors, however, represent consistent biases that affect all measurements in the same way and cannot be reduced by simply taking more observations. Option D correctly separates these error types:

[Fe/H] = -0.50 \pm 0.05_{stat} \pm 0.10_{sys}

. This notation clearly communicates both the precision of the measurement (statistical uncertainty) and the potential bias from the atmospheric model (systematic uncertainty), allowing other astronomers to properly interpret and use the result. Option A incorrectly adds the errors linearly (0.05 + 0.10 = 0.15), which overestimates the total uncertainty and obscures the different error sources. Option B shifts the central value but ignores the systematic uncertainty entirely, which is misleading since the bias might not actually be present. Option C arbitrarily averages the errors, which has no statistical justification and loses important information about error sources. Remember: Always report statistical and systematic uncertainties separately in scientific measurements. This transparency helps the scientific community properly evaluate and build upon your work.

Question 2

An astronomer measures the visual magnitude of an asteroid on five consecutive nights, obtaining the values: 15.2, 15.4, 15.3, 15.8, 15.3. They calculate the mean as 15.4 and the standard error on the mean as 0.1. A second astronomer suggests the asteroid might be a variable object. Which statement provides the most scientifically valid reason to question the use of the standard error on the mean as the final uncertainty?

The standard error on the mean only reflects the measurement's random error, not any potential intrinsic variability of the asteroid. (correct answer)
The true uncertainty can only be known if the asteroid's true magnitude is known.
The uncertainty should be the full range of the data, which is 15.8 - 15.2 = 0.6.
The sample size of five measurements is too small to calculate a meaningful mean or standard error.

Explanation: When analyzing astronomical measurements, you need to distinguish between two types of uncertainty: measurement error (how precisely you can read your instrument) and intrinsic variability (whether the object itself is actually changing). This distinction is crucial for interpreting your results correctly. The standard error on the mean reflects only the scatter in your measurements - it tells you how precisely you've measured whatever the asteroid was doing during those five nights. However, if the asteroid is genuinely variable (changing brightness over time), then this scatter isn't just measurement noise - it's real physical variation. In that case, using standard error as your final uncertainty would be misleading because you'd be claiming much higher precision than you actually have. Looking at the options: Answer A correctly identifies that standard error only captures random measurement error, not potential intrinsic variability of the source. Answer B is wrong because you don't need to know the "true" magnitude to assess uncertainty properly - uncertainty analysis is about quantifying what you don't know. Answer C incorrectly suggests using the full range as uncertainty, which isn't statistically sound and would overestimate uncertainty even for a variable object. Answer D is incorrect because five measurements are sufficient to calculate meaningful statistics, though more would be better for detecting variability. The key takeaway: When you suspect astronomical variability, your uncertainty analysis must account for both measurement precision AND the intrinsic scatter of the object. Standard error alone only addresses the first component, potentially giving false confidence in your measurement precision.

Question 3

Two astronomical discoveries are announced:

Discovery 1: A star's brightness is found to dim by $0.010 \pm 0.001$ percent during a planetary transit.
Discovery 2: A galaxy is found to have a recession velocity of $5000 \pm 1000$ km/s.

Which statement provides the most astute comparison of these two results?

The galaxy velocity is a more certain measurement because its absolute value is much larger than the transit depth.
The transit measurement is more statistically significant because its signal-to-noise ratio is higher. (correct answer)
Both measurements are equally reliable because their relative uncertainties are both less than 25%.
The galaxy velocity measurement is more precise because its absolute error is larger.

Explanation: Statistical significance is best assessed by the signal-to-noise ratio (S/N = central value / uncertainty). For Discovery 1, S/N = 0.010 / 0.001 = 10 (a 10-sigma detection). For Discovery 2, S/N = 5000 / 1000 = 5 (a 5-sigma detection). Since 10 > 5, the transit measurement is more statistically significant. This demonstrates that a very small physical effect can be measured with very high confidence.

Question 4

The observed precession of Mercury's perihelion is 574.1 arcseconds/century. Newtonian physics predicts a precession of 531.6 arcseconds/century. An early 20th-century astronomer, Researcher X, measures the excess precession (the part not explained by Newton) to be $41.9 \pm 2.1$ arcseconds/century. Einstein's General Relativity (GR) predicts this excess to be 42.9 arcseconds/century. Which statement accurately assesses Researcher X's result?

The measurement is inconclusive because its uncertainty (2.1) is large compared to the small difference between the measurement (41.9) and the GR prediction (42.9).
The measurement refutes General Relativity because its central value of 41.9 does not exactly match the predicted value of 42.9.
The measurement strongly refutes the Newtonian prediction of zero excess and is statistically consistent with the prediction of General Relativity. (correct answer)
The measurement is flawed because the true excess precession (42.5) is not the central value of the measurement's range.

Explanation: The key scientific questions are whether the measurement is consistent with GR and inconsistent with Newton. The Newtonian prediction is an excess of 0. Researcher X's result of

41.9 \pm 2.1

excludes 0 by about 20 sigma (41.9/2.1), strongly refuting it. The GR prediction is 42.9. Researcher X's 1-sigma range is [39.8, 44.0]. Since 42.9 falls within this range, the measurement is statistically consistent with GR. Therefore, the result supports GR and refutes Newton.

Question 5

Two independent teams measure the Hubble Constant. Team A reports $H_0 = 73.2 \pm 1.3$ km/s/Mpc, while Team B reports $H_0 = 67.8 \pm 0.9$ km/s/Mpc. What is the most scientifically sound interpretation of these results?

The measurements are consistent, and the true value likely lies in the range where their error bars overlap.
Team B's result is more accurate than Team A's because its reported uncertainty is smaller.
The results are statistically inconsistent, suggesting the presence of unaccounted-for systematic errors in one or both experiments. (correct answer)
A weighted average of the two results should be calculated to find the most likely true value for the Hubble Constant.

Explanation: The 1-sigma range for Team A is [71.9, 74.5] km/s/Mpc, and for Team B it is [66.9, 68.7] km/s/Mpc. These ranges do not overlap, indicating a statistically significant disagreement (often called 'tension'). Such a disagreement between precise measurements suggests that one or both experiments may have systematic errors that were not accounted for in their uncertainty budgets. Averaging inconsistent results is not appropriate, and smaller uncertainty relates to precision, not necessarily accuracy.

Question 6

An astronomer uses a spectrograph to measure the radial velocities of stars. They notice that all their measurements for stars with known, stable velocities are consistently about 5 km/s higher than the established catalog values. This discrepancy is most likely due to:

random statistical error caused by photon noise in the stellar spectra.
a systematic error from incorrect wavelength calibration of the spectrograph. (correct answer)
human error in recording the data for each star individually.
the intrinsic variability of the standard stars used for comparison.

Explanation: A consistent offset that affects all measurements in the same direction (e.g., all are 5 km/s too high) is the signature of a systematic error. Random errors (like photon noise or individual recording mistakes) would cause scatter, with some measurements being too high and others too low. An incorrect wavelength calibration in a spectrograph would shift the entire spectrum, leading to a systematic offset in all calculated velocities.

Question 7

An attempt to measure the trigonometric parallax of a very distant quasar yields a result of $0.0005 \pm 0.0008$ arcseconds. For all practical purposes, quasars are too distant to have a measurable parallax (the true value is effectively zero). Which conclusion is supported by this measurement?

The measurement demonstrates that the quasar has a small but non-zero parallax of exactly 0.0005 arcseconds.
The experiment failed because the uncertainty is larger than the measured central value, making the result meaningless.
The measurement is consistent with the quasar having zero parallax and establishes an upper limit on its possible value. (correct answer)
The measurement proves that the parallax of the quasar is exactly zero, confirming theoretical expectations.

Explanation: The range of plausible values given by the measurement is [0.0005 - 0.0008, 0.0005 + 0.0008], or [-0.0003, 0.0013] arcseconds. Since this range comfortably includes zero, the result is statistically consistent with the expected null result (zero parallax). This is not a failed experiment; rather, it successfully shows that any parallax is too small to be detected and places an upper limit on its value. A measurement can never prove a value is exactly zero.

Question 8

Three students measure the period of a Cepheid variable star:

Student 1: $5.36 \pm 0.10$ days
Student 2: $5.45 \pm 0.02$ days
Student 3: $5.41 \pm 0.08$ days

To obtain the single best estimate for the period from this combined dataset, how should these measurements be treated?

A simple arithmetic mean should be calculated, as all three measured the same star under similar conditions.
A weighted average should be calculated, giving the most weight to Student 2's measurement. (correct answer)
Student 2's measurement should be used alone as the final result, since it is the most precise.
The median value of the three central values should be used to minimize the effect of any potential outliers.

Explanation: When combining multiple measurements of the same quantity that have different uncertainties, the statistically optimal method is to calculate a weighted average. The weight given to each measurement is typically inversely proportional to its variance (the square of its uncertainty). Therefore, the measurement with the smallest uncertainty (Student 2) receives the highest weight, as it provides the most information.

Question 9

A student measures the distance to a galaxy to be $d = 50 \pm 15$ Mpc. A published value, based on a more sophisticated method, is $d_{pub} = 62$ Mpc. The student claims their result is 'consistent with the published value'. Why is this claim justified?

Because the published value of 62 Mpc lies within the student's 1-sigma uncertainty range. (correct answer)
Because the student's central value of 50 Mpc is reasonably close to the published value of 62 Mpc.
Because the student's measurement has a large uncertainty, indicating that almost any value is possible.
Because the percent difference between the two values is less than the student's percent uncertainty.

Explanation: When interpreting scientific measurements with uncertainties, the key question is whether two values are statistically consistent—meaning their ranges of possible values overlap when accounting for measurement errors. The student's measurement is

d = 50 \pm 15

Mpc, which means the true distance likely falls between 35 Mpc and 65 Mpc (the 1-sigma range). Since the published value of 62 Mpc falls within this range, the measurements are indeed consistent. This is the fundamental principle behind comparing experimental results: if the uncertainty ranges overlap, the measurements don't contradict each other. Looking at the wrong answers: Option B is incorrect because "reasonably close" is subjective and unscientific—what matters is whether values fall within uncertainty ranges, not our intuitive sense of closeness. Option C misunderstands what large uncertainty means; while the student's measurement is imprecise, this doesn't make "almost any value possible"—the uncertainty still defines meaningful bounds. Option D introduces an irrelevant comparison method; percent differences aren't the standard way to assess consistency in astronomy measurements. The correct answer is A because it correctly identifies that statistical consistency requires the published value to fall within the student's uncertainty range, which it does (62 Mpc lies between 35-65 Mpc). Study tip: When comparing astronomical measurements, always check if values fall within each other's uncertainty ranges rather than just comparing central values. This is the standard criterion for determining if measurements are consistent across all physical sciences.

Question 10

An astronomer measures the mass of an exoplanet and reports the result as $1.50 \pm 0.08 M_{Jup}$ , where the uncertainty represents one standard deviation (1-sigma). Assuming the measurement errors are normally distributed, what does this reported uncertainty imply?

The true mass of the exoplanet is guaranteed to be between $1.42$ and $1.58 M_{Jup}$ .
If the same measurement were repeated many times, approximately 68% of the results would fall between $1.42$ and $1.58 M_{Jup}$ . (correct answer)
All individual measurements taken by the astronomer in their experiment fell between $1.42$ and $1.58 M_{Jup}$ .
There is a 95% probability that the true mass of the exoplanet is exactly $1.50 M_{Jup}$ .

Explanation: For a normal (Gaussian) distribution of errors, one standard deviation (1-sigma) defines the range around the mean that is expected to contain approximately 68.3% of all subsequent measurements. It is a probabilistic statement about the spread of the data, not a guarantee about the location of the true value. The true value itself has a 68% probability of being in the [1.42, 1.58] range, but it is not guaranteed.

Question 11

The measurement demonstrates that the quasar has a small but non-zero parallax of exactly 0.0005 arcseconds.
The experiment failed because the uncertainty is larger than the measured central value, making the result meaningless.
The measurement is consistent with the quasar having zero parallax and establishes an upper limit on its possible value. (correct answer)
The measurement proves that the parallax of the quasar is exactly zero, confirming theoretical expectations.

Question 12

An astronomer measures a star's parallax as $p = 0.050 \pm 0.002$ arcseconds. A later, highly reliable measurement from a space telescope reveals the true parallax to be $0.061$ arcseconds. How should the astronomer's original measurement be characterized?

It had high precision but low accuracy. (correct answer)
It had high accuracy but low precision.
It suffered primarily from large random statistical errors.
It was a valid measurement because the difference from the true value is physically small.

Explanation: Precision refers to the size of the uncertainty relative to the measurement. An uncertainty of 0.002 on a value of 0.050 (a 4% uncertainty) indicates high precision. Accuracy refers to how close the measurement is to the true value. The measured range is [0.048, 0.052] arcseconds. The true value of 0.061 is far outside this range, indicating low accuracy. This discrepancy points to a significant systematic error, not just random statistical errors.

Question 13

An astronomical survey pipeline reports a potential signal from a distant galaxy with a flux measurement of $F = 1.2 \pm 0.4$ arbitrary units, where the background noise level is assumed to be zero. What is the most appropriate interpretation of this result?

The signal is not real because the error bar (0.4) is a large fraction (one-third) of the central value (1.2).
The signal is a definite discovery because its error bar does not include zero, proving it is not a background fluctuation.
The signal is likely a random fluctuation of the background noise because the signal-to-noise ratio is low.
The result is a 3-sigma detection, which is typically considered evidence but may require more data for a definitive claim. (correct answer)

Explanation: The significance of a detection is measured by the signal-to-noise ratio, which is the central value divided by the uncertainty. Here,

1.2 / 0.4 = 3

. This is known as a '3-sigma' detection. In many fields of science, a 3-sigma result is considered evidence for a signal, but a higher threshold (like 5-sigma) is often required for a formal 'discovery' claim, as a 3-sigma event can still occur by chance about 0.3% of the time.

Question 14

It had high precision but low accuracy. (correct answer)
It had high accuracy but low precision.
It suffered primarily from large random statistical errors.
It was a valid measurement because the difference from the true value is physically small.

Question 15

The kinetic energy $E$ of a meteor is given by $E = \frac{1}{2}mv^2$ . If a scientist measures the meteor's mass $m$ with a 4% uncertainty and its velocity $v$ with a 3% uncertainty, what is the approximate percent uncertainty in the calculated kinetic energy $E$ ?

Exactly 7%, the simple sum of the individual uncertainties.
Exactly 10%, the sum of the mass uncertainty and twice the velocity uncertainty.
Approximately 7.2%, derived from combining the uncertainties in quadrature with appropriate weighting for exponents. (correct answer)
Approximately 5%, the quadrature sum of the raw uncertainties, ignoring the exponent on velocity.

Explanation: For a formula

E \propto m^a v^b

, the fractional uncertainties add in quadrature, with each term multiplied by its exponent:

(\Delta E/E)^2 = (a \cdot \Delta m/m)^2 + (b \cdot \Delta v/v)^2

. Here, a=1 and b=2. The percent uncertainty is

\sqrt{(1 \cdot 4\%)^2 + (2 \cdot 3\%)^2} = \sqrt{4^2 + 6^2} = \sqrt{16 + 36} = \sqrt{52} \approx 7.2\%

. Simply adding the uncertainties (linearly) is a common misconception.

Question 16

The measurement is inconclusive because its uncertainty (2.1) is large compared to the small difference between the measurement (41.9) and the GR prediction (42.9).
The measurement refutes General Relativity because its central value of 41.9 does not exactly match the predicted value of 42.9.
The measurement strongly refutes the Newtonian prediction of zero excess and is statistically consistent with the prediction of General Relativity. (correct answer)
The measurement is flawed because the true excess precession (42.5) is not the central value of the measurement's range.

Explanation: The key scientific questions are whether the measurement is consistent with GR and inconsistent with Newton. The Newtonian prediction is an excess of 0. Researcher X's result of

41.9 \pm 2.1

Question 17

The true mass of the exoplanet is guaranteed to be between $1.42$ and $1.58 M_{Jup}$ .
If the same measurement were repeated many times, approximately 68% of the results would fall between $1.42$ and $1.58 M_{Jup}$ . (correct answer)
All individual measurements taken by the astronomer in their experiment fell between $1.42$ and $1.58 M_{Jup}$ .
There is a 95% probability that the true mass of the exoplanet is exactly $1.50 M_{Jup}$ .

Question 18

Combine the errors to report $[Fe/H] = -0.50 \pm 0.15$ .
Adjust the central value and report $[Fe/H] = -0.60 \pm 0.05$ .
Average the two errors and report $[Fe/H] = -0.50 \pm 0.075$ .
Report the result as $[Fe/H] = -0.50 \pm 0.05_{stat} \pm 0.10_{sys}$ . (correct answer)

[Fe/H] = -0.50 \pm 0.05_{stat} \pm 0.10_{sys}

Question 19

An astronomer is fitting a straight line, $y = mx + b$ , to a set of data points that have uncertainties in the y-direction. The fitting procedure being used is a 'chi-squared minimization', which aims to find the line that minimizes the value of $\chi^2 = \sum \frac{(y_i - (mx_i+b))^2}{\sigma_i^2}$ , where $y_i$ are the data points and $\sigma_i$ are their corresponding uncertainties.

Based on the formula for $\chi^2$ , how do the uncertainties $\sigma_i$ of the data points influence the best-fit line?

The line will be pulled preferentially toward data points with the smallest uncertainties, as they contribute most to the $\chi^2$ value if the line is far from them. (correct answer)
The line will be pulled preferentially toward data points with the largest uncertainties, as they have the largest residuals.
The uncertainties do not influence the slope or intercept of the line, but only determine the uncertainty in the final fitted parameters $m$ and $b$ .
The line will pass exactly through the two data points with the smallest and largest uncertainties, as these anchor the fit.

Explanation: When you encounter chi-squared fitting problems, focus on how the weighting in the formula affects which data points have the most influence on the final result. In the chi-squared formula

\chi^2 = \sum \frac{(y_i - (mx_i+b))^2}{\sigma_i^2}

, each data point's contribution is divided by the square of its uncertainty. This creates a weighting system where points with smaller uncertainties (smaller

\sigma_i

) have larger denominators when their residuals are large, making them contribute more heavily to the total

\chi^2

value. Since the fitting procedure minimizes

\chi^2

, the algorithm works hardest to keep the line close to these high-precision points. This makes choice A correct—the line is pulled toward points with the smallest uncertainties. Choice B incorrectly suggests the line favors points with large uncertainties. In reality, points with large

\sigma_i

values have their contributions diminished by the large denominator, giving them less influence. Choice C misunderstands the weighting entirely—uncertainties absolutely do affect the fitted parameters

m

and

b

, not just their final uncertainties. Choice D describes a completely different fitting method; chi-squared minimization doesn't force the line through specific anchor points. Remember this pattern: in weighted fitting procedures, look at what's in the denominator of the weighting term. Smaller denominators mean greater influence. Data points you trust more (smaller uncertainties) should logically have more say in determining the best-fit line.

Question 20

The cosmic microwave background (CMB) has a nearly uniform temperature across the sky. Two different experiments measure the CMB temperature in the same patch of sky. Experiment A reports $T_A = 2.7250 \pm 0.0020$ K. Experiment B reports $T_B = 2.7285 \pm 0.0015$ K. Are these two measurements statistically consistent?

Yes, because the difference in their central values is only 0.0035 K, which is a very small number.
No, because the more precise measurement from Experiment B does not have its central value within the error range of Experiment A.
Yes, because both measurements are consistent with the established CMB temperature of approximately 2.73 K.
No, because their 1-sigma error bars do not overlap with each other. (correct answer)

Explanation: To check for consistency, we examine the ranges defined by the 1-sigma uncertainties. The range for Experiment A is [2.7230, 2.7270] K. The range for Experiment B is [2.7270, 2.7300] K. These two ranges only touch at a single point (2.7270 K) and do not overlap. This indicates a statistical tension or inconsistency between the results. For two measurements to be considered consistent, their error bars must overlap. The absolute smallness of the difference is irrelevant without considering the size of the uncertainties.