TRB PG Assistant - Design of experiments - principles of experimentation, analysis of variance, completely randomized design, randomized block design, Latin square design Online Test (MCQ Quiz)

1. What is the primary goal of the Design of Experiments (DOE)?

A) To collect as much data as possible, regardless of structure. B) To systematically plan and conduct experiments to efficiently obtain valid and objective answers to the questions being asked. C) To analyze existing data without conducting new experiments. D) To develop theoretical statistical models without experimental validation.

2. Which of the following is NOT a fundamental principle of experimentation?

A) Replication B) Randomization C) Blocking D) Extrapolation

3. What does the principle of Replication in experimental design refer to?

A) Using the same experimental unit for multiple treatments. B) Repeating an experiment under identical conditions to assess variability. C) Applying different treatments to the same experimental unit. D) Ensuring all experimental units are identical.

4. The purpose of Randomization in experimental design is to:

A) Ensure that treatments are applied in a predetermined order. B) Minimize the effect of unknown or uncontrolled variables and avoid bias. C) Allow for the grouping of similar experimental units. D) Increase the number of experimental units.

5. What is the main advantage of using Blocking in experimental design?

A) To increase the number of treatments. B) To reduce the variability within blocks, making it easier to detect treatment differences. C) To eliminate the need for randomization. D) To simplify the experimental setup.

6. Analysis of Variance (ANOVA) is a statistical technique used to:

A) Determine the correlation between two variables. B) Compare the means of three or more groups by analyzing variances. C) Calculate the probability of a specific event occurring. D) Summarize data using measures of central tendency.

7. In ANOVA, the null hypothesis ($H_0$) typically states that:

A) All group means are significantly different. B) At least one group mean is different from the others. C) All group means are equal. D) The variance between groups is zero.

8. The F-statistic in ANOVA is calculated as the ratio of:

A) Mean Square Error (MSE) to Mean Square Treatment (MST). B) Mean Square Treatment (MST) to Mean Square Error (MSE). C) Total Sum of Squares (SST) to Mean Square Treatment (MST). D) Mean Square Treatment (MST) to Total Sum of Squares (SST).

9. A 'treatment' in the context of experimental design refers to:

A) A specific experimental unit. B) A condition or factor applied to the experimental units whose effect is being studied. C) The overall experimental setup. D) The error associated with measurements.

10. A 'block' in experimental design is used to:

A) Group experimental units that are similar in characteristics that might affect the response. B) Represent a specific treatment level. C) Isolate the effect of a nuisance variable. D) Both B and C

11. Which experimental design is the simplest and most basic, involving the random assignment of treatments to experimental units without any grouping?

A) Randomized Block Design (RBD) B) Latin Square Design (LSD) C) Completely Randomized Design (CRD) D) Factorial Design

12. In a Completely Randomized Design (CRD), what is the primary source of variation considered?

A) Differences between blocks. B) Differences between treatments and random error. C) Interactions between factors. D) Variations due to rows and columns.

13. When would a Randomized Block Design (RBD) be more appropriate than a CRD?

A) When there is no known source of variability among experimental units. B) When there is a known source of variability (e.g., location, time) that can be used to form blocks. C) When only two treatments are being compared. D) When all experimental units are completely homogeneous.

14. In an RBD, the total variation is partitioned into variation due to:

A) Treatments, blocks, and error. B) Treatments and error only. C) Blocks and error only. D) Rows, columns, and treatments.

15. The purpose of blocking in RBD is to:

A) Increase the number of experimental units. B) Reduce the experimental error by accounting for variability within blocks. C) Introduce bias into the experiment. D) Allow for more treatments to be tested simultaneously.

16. What is a key characteristic of the Latin Square Design (LSD)?

A) It involves blocking in only one direction. B) It controls for variation in two different directions simultaneously. C) It is suitable for any number of treatments. D) It requires a large number of experimental units.

17. In a Latin Square Design, the number of treatment levels must equal:

A) The number of rows plus the number of columns. B) The number of rows or the number of columns. C) The number of experimental units. D) The total sum of squares.

18. What are the two sources of variation, besides treatments, that LSD attempts to control?

A) Time and location. B) Rows and columns. C) Operators and machines. D) Both B and C

19. A potential drawback of the Latin Square Design is:

A) It requires too few experimental units. B) It can only be used for a small number of treatments. C) It assumes no interaction between the blocking factors and the treatments. D) It is difficult to analyze statistically.

20. In ANOVA, the Sum of Squares Total (SST) represents:

A) The variation explained by the treatments. B) The variation explained by the blocks. C) The total variation in the data around the grand mean. D) The unexplained variation (error).

21. The Sum of Squares Treatment (SSTr) in ANOVA measures:

A) The variation among the block means. B) The variation among the treatment means. C) The total variation in the dataset. D) The random error variation.

22. The Sum of Squares Error (SSE) represents:

A) The variation explained by treatments and blocks. B) The variation due to systematic differences between blocks. C) The variation that cannot be explained by treatments or blocks, representing random error. D) The total variation in the data.

23. Degrees of freedom for treatments in a CRD with 'a' treatments is:

A) a - 1 B) N - a C) N - 1 D) a

24. Degrees of freedom for error in a CRD with N total observations and 'a' treatments is:

A) a - 1 B) N - a C) N - 1 D) a - 1

25. Which design is most efficient when the experimental units are expected to be homogeneous?

A) Completely Randomized Design (CRD) B) Randomized Block Design (RBD) C) Latin Square Design (LSD) D) Split-Plot Design

26. If an experiment involves studying the effect of fertilizer type (Factor A) and watering frequency (Factor B) on plant growth, which design would be most suitable if you also want to control for variations due to different soil types?

A) CRD B) RBD with soil type as blocks C) LSD with soil type as one blocking factor D) Factorial design within an RBD or LSD framework

27. The 'grand mean' in ANOVA is:

A) The average of the treatment means. B) The average of the block means. C) The average of all observations in the experiment. D) The average of the error terms.

28. In a CRD, if the F-calculated value is greater than the F-critical value (at a given significance level), we:

A) Accept the null hypothesis. B) Reject the null hypothesis. C) Conclude there is no significant difference. D) Increase the sample size.

29. What is the primary assumption of ANOVA regarding the error terms?

A) They are independent, normally distributed with mean zero and equal variances. B) They are dependent and non-normally distributed. C) They have unequal variances across treatments. D) They are correlated with the treatment effects.

30. Which of the following is a method for post-hoc analysis after a significant ANOVA result?

A) Tukey's HSD (Honestly Significant Difference) B) Chi-Squared Test C) Regression Analysis D) Correlation Coefficient

31. The purpose of post-hoc tests in ANOVA is to:

A) Determine if the overall model is significant. B) Identify which specific group means differ significantly from each other. C) Control for blocking effects. D) Calculate the F-statistic.

32. In an RBD, if the blocking factor is highly effective, we expect:

A) The Sum of Squares for Blocks (SSB) to be large and the error variance to be small. B) The Sum of Squares for Treatments (SSTr) to be small. C) The F-statistic for treatments to be non-significant. D) The error variance to increase.

33. When is a Latin Square Design particularly useful?

A) When there are many treatments and few experimental units. B) When there are three or more potential sources of variation to control. C) When there is only one source of variation to control. D) When the experimental units are completely homogeneous.

34. Consider an experiment to test 5 different oven temperatures (treatments) for baking cakes. If you suspect variations due to the day of the week and the baker, which design would be most appropriate?

A) CRD B) RBD (blocking by day or baker) C) LSD (rows=day, columns=baker, cells=temperature) D) Factorial Design

35. The 'experimental unit' is defined as:

A) A specific treatment level. B) The smallest unit to which a treatment is applied independently. C) A block of similar units. D) A single observation.

36. In ANOVA, the Mean Square Treatment (MST) is calculated as:

A) SSTr / (N - a) B) SSTr / (a - 1) C) SSE / (N - 1) D) SST / (a - 1)

37. The Mean Square Error (MSE) in ANOVA is calculated as:

A) SSE / (a - 1) B) SSE / (N - a) C) SSTr / (N - 1) D) SSE / (N - 1)

38. What is the primary limitation of CRD when there's heterogeneity in experimental units?

A) It may lead to higher experimental error and reduced power to detect treatment differences. B) It requires too many treatments. C) It is difficult to analyze. D) It cannot handle more than two treatments.

39. In an RBD with 'r' blocks and 'a' treatments, the degrees of freedom for error are:

A) (r-1)(a-1) B) ra - 1 C) r + a - 1 D) ra - r - a + 1

40. If the F-test for treatments in an RBD is not significant, what can be concluded?

A) All treatment means are significantly different. B) There is no statistically significant difference between the treatment means, after accounting for block effects. C) The blocking was ineffective. D) The experiment should be repeated with more blocks.

41. The assumption of 'no interaction' between treatments and blocking factors is crucial for which design?

A) CRD B) RBD C) LSD D) All of the above

42. What statistical test is commonly used to check the assumption of normality of residuals in ANOVA?

A) Durbin-Watson test B) Shapiro-Wilk test C) Levene's test D) Bartlett's test

43. Which test is used to check the assumption of homogeneity of variances (homoscedasticity) in ANOVA?

A) Anderson-Darling test B) Kolmogorov-Smirnov test C) Levene's test D) Runs test

44. If the assumption of equal variances is violated, which alternative approach might be considered?

A) Using a CRD instead of RBD. B) Performing Welch's ANOVA or using transformations. C) Increasing the number of blocks. D) Ignoring the violation and proceeding with standard ANOVA.

45. What is the main advantage of using a factorial experiment within a designed structure like CRD, RBD, or LSD?

A) It simplifies the analysis. B) It allows for the study of main effects and interaction effects between factors. C) It requires fewer experimental units. D) It eliminates the need for randomization.

46. In the context of DOE, 'factor' refers to:

A) A specific outcome measured. B) A variable whose level is manipulated or set by the experimenter. C) A random error component. D) A blocking variable.

47. What is the purpose of 'random error' or 'residual' in ANOVA?

A) To represent the systematic differences between treatments. B) To represent the variation due to known blocking factors. C) To represent the unexplained variation in the response variable after accounting for treatment and blocking effects. D) To represent the overall average effect.

48. Which type of error is controlled by blocking?

A) Random error. B) Systematic error due to known sources of variation. C) Experimental error. D) Sampling error.

Design of experiments - principles of experimentation, analysis of variance, completely randomized design, randomized block design, Latin square design - Online Test

Test Results