Biostatistics 5 QA Exam Training Questions and Answers PDF January 2025

Biostatistics 5 QA Exams Training January 2025 Marina NIKOLIDAKI  Question What does GPT stand for in ChatGPT? A) Generative Pre-trained Transformer B) Generalized Processing Tool C) Graphical Programming T...

Biostatistics 5 QA Exams Training January 2025 Marina NIKOLIDAKI  Question What does GPT stand for in ChatGPT? A) Generative Pre-trained Transformer B) Generalized Processing Tool C) Graphical Programming Technique D) Generalized Prompt Translator January 2025 Marina NIKOLIDAKI  Answer  A) Generative Pre-trained Transformer January 2025 Marina NIKOLIDAKI  Question Which library is commonly used to deploy ChatGPT models in Python? A) TensorFlow B) OpenAI API C) PyTorch D) NumPy January 2025 Marina NIKOLIDAKI  Answer  B) OpenAI API January 2025 Marina NIKOLIDAKI  Question The assumption of linearity ensures: A) Predictor and response variables have a linear relationship B) Residuals are normally distributed C) Predictors are independent of one another D) Variance is constant across observations January 2025 Marina NIKOLIDAKI   A) Predictor and response variables have a linear Answer relationship January 2025 Marina NIKOLIDAKI  How does ChatGPT assist in summarizing biostatistics papers? A) It automatically validates statistical results Question B) It creates concise summaries based on provided abstracts C) It extracts graphs and charts from the paper D) It runs the code directly for analysis January 2025 Marina NIKOLIDAKI   B) It creates concise summaries based on Answer provided abstracts January 2025 Marina NIKOLIDAKI  Question What type of errors are common when using ChatGPT for generating statistical code? A) Syntax errors B) Contextual misunderstandings C) Hardware dependencies D) Mathematical calculation errors January 2025 Marina NIKOLIDAKI  Answer  B) Contextual misunderstandings January 2025 Marina NIKOLIDAKI  Question What is a key ethical consideration when using ChatGPT in academia? A) Avoiding sensitive prompts B) Properly acknowledging AI assistance C) Limiting the use of advanced models D) Testing generated code for performance January 2025 Marina NIKOLIDAKI Answer  B) Properly acknowledging AI assistance January 2025 Marina NIKOLIDAKI  What does a well-written ChatGPT prompt achieve? A) Reduces model overfitting Question B) Avoids errors and vague responses C) Optimizes neural network performance D) Guarantees accurate citations January 2025 Marina NIKOLIDAKI Answer  B) Avoids errors and vague responses January 2025 Marina NIKOLIDAKI  The Durbin-Watson test is used to check for autocorrelation in the residuals of a regression model. A) A value around 2 indicates no autocorrelation Question B) A value closer to 0 suggests negative autocorrelation C) A value closer to 4 indicates positive autocorrelation D) It evaluates homoscedasticity January 2025 Marina NIKOLIDAKI  Answer  A) A value around 2 indicates no autocorrelation January 2025 Marina NIKOLIDAKI  What is the main advantage of using ChatGPT for coding tasks? A) It generates error-free code Question B) It provides rapid prototyping and suggestions C) It eliminates the need for human expertise D) It guarantees performance optimization January 2025 Marina NIKOLIDAKI Answer  B) It provides rapid prototyping and suggestions January 2025 Marina NIKOLIDAKI  Which task is ChatGPT least effective at? A) Debugging simple R code Question B) Designing complete biostatistical studies C) Summarizing lengthy texts D) Generating sample datasets January 2025 Marina NIKOLIDAKI Answer  B) Designing complete biostatistical studies January 2025 Marina NIKOLIDAKI  Question How can you validate code generated by ChatGPT? A) Run it as-is without testing B) Cross-check against documentation and test in software C) Assume it’s error-free if it runs D) Use the ChatGPT debugging mode January 2025 Marina NIKOLIDAKI   B) Cross-check against documentation and test in Answer software January 2025 Marina NIKOLIDAKI  What is the main difference between GPT and BERT? A) GPT is generative, while BERT is analytical Question B) GPT focuses on coding, while BERT focuses on text generation C) GPT is bidirectional, while BERT is unidirectional D) GPT is for summarization, while BERT is for classification January 2025 Marina NIKOLIDAKI Answer  A) GPT is generative, while BERT is analytical January 2025 Marina NIKOLIDAKI  Question What does the term 'prompt engineering' refer to? A) Adjusting GPT's internal algorithms B) Designing effective input questions to guide AI responses C) Training GPT models from scratch D) Debugging ChatGPT for advanced tasks January 2025 Marina NIKOLIDAKI   B) Designing effective input questions to guide AI Answer responses January 2025 Marina NIKOLIDAKI  Why are iterative prompts recommended when using ChatGPT? A) They improve tokenization speed Question B) They refine and focus AI-generated outputs C) They train the model faster D) They automatically summarize large datasets January 2025 Marina NIKOLIDAKI Answer  B) They refine and focus AI-generated outputs January 2025 Marina NIKOLIDAKI  Question What type of language does ChatGPT-generated text often lack? A) Informal language B) Domain-specific terminology C) Contextual relevance D) Nuanced critical arguments January 2025 Marina NIKOLIDAKI Answer  D) Nuanced critical arguments January 2025 Marina NIKOLIDAKI  What should be avoided when summarizing a paper using ChatGPT? Question A) Providing clear instructions B) Asking for focus on methods and results C) Assuming the summary is 100% accurate D) Comparing ChatGPT outputs with the original text January 2025 Marina NIKOLIDAKI Answer  C) Assuming the summary is 100% accurate January 2025 Marina NIKOLIDAKI  Question What role can ChatGPT play in experimental design? A) Developing complete experimental protocols B) Suggesting ideas for study designs C) Verifying statistical assumptions automatically D) Avoiding human review of the design January 2025 Marina NIKOLIDAKI Answer  B) Suggesting ideas for study designs January 2025 Marina NIKOLIDAKI  Question Dunnett’s test is used to compare: A) Multiple treatments to each other B) Multiple treatments to a single control group C) Treatment groups based on non-parametric criteria D) Mean variances across groups January 2025 Marina NIKOLIDAKI Answer  B) Multiple treatments to a single control group January 2025 Marina NIKOLIDAKI  The role of a covariate in ANCOVA is to: A) Adjust for categorical group effects Question B) Control for variability in a continuous variable C) Replace independent predictors in the model D) Test for interaction effects January 2025 Marina NIKOLIDAKI Answer  B) Control for variability in a continuous variable January 2025 Marina NIKOLIDAKI  The assumption of homogeneity of regression slopes in ANCOVA ensures: A) Identical mean values across groups Question B) Consistent relationships between the covariate and dependent variable across groups C) Normality of residuals across different treatment groups D) Independence of observations January 2025 Marina NIKOLIDAKI   B) Consistent relationships between the covariate Answer and dependent variable across groups January 2025 Marina NIKOLIDAKI  Question Cross-validation is useful because it: A) Splits data to evaluate model generalizability B) Reduces bias by using fewer predictors C) Tests residuals for normality and homoscedasticity D) Simplifies model comparison January 2025 Marina NIKOLIDAKI Answer  A) Splits data to evaluate model generalizability January 2025 Marina NIKOLIDAKI  LASSO regression differs from traditional regression by: A) Increasing coefficients to improve accuracy Question B) Shrinking coefficients to zero for feature selection C) Using quadratic penalties to avoid multicollinearity D) Transforming predictors into uncorrelated components January 2025 Marina NIKOLIDAKI   B) Shrinking coefficients to zero for feature Answer selection January 2025 Marina NIKOLIDAKI  When comparing regression models, a lower AIC value: Question A) Indicates a more complex but less accurate model B) Suggests the model is overfitting C) Balances goodness-of-fit and model complexity D) Shows independence of residuals January 2025 Marina NIKOLIDAKI Answer  C) Balances goodness-of-fit and model complexity January 2025 Marina NIKOLIDAKI  Question Random effects in mixed-effects models account for: A) Fixed treatment-level effects B) Variability specific to individual subjects or clusters C) Residual errors unrelated to predictors D) Interaction effects among predictors January 2025 Marina NIKOLIDAKI   B) Variability specific to individual subjects or Answer clusters January 2025 Marina NIKOLIDAKI  PCR is advantageous in datasets with high multicollinearity because it: A) Uses uncorrelated principal components as Question predictors B) Identifies irrelevant variables and eliminates them C) Reduces overfitting by increasing residual variance D) Balances variance and bias optimally January 2025 Marina NIKOLIDAKI   A) Uses uncorrelated principal components as Answer predictors January 2025 Marina NIKOLIDAKI  Question Scheffé’s method is ideal for: A) Comparing treatment means to a control B) Exploring all possible contrasts in group means C) Testing for interaction effects in regression models D) Analyzing within-group variability January 2025 Marina NIKOLIDAKI Answer  B) Exploring all possible contrasts in group means January 2025 Marina NIKOLIDAKI  Homoscedasticity in regression refers to: A) The absence of autocorrelation in residuals Question B) Constant variance of residuals across predictor levels C) The linearity between predictors and response variables D) Normality of residuals January 2025 Marina NIKOLIDAKI   B) Constant variance of residuals across predictor Answer levels January 2025 Marina NIKOLIDAKI  Question Independence of residuals is critical because: A) Dependent residuals can inflate Type I errors B) It ensures multicollinearity is minimized C) It guarantees constant variance across observations D) It eliminates outliers January 2025 Marina NIKOLIDAKI Answer  A) Dependent residuals can inflate Type I errors January 2025 Marina NIKOLIDAKI  Question Q-Q plots help evaluate: A) Homoscedasticity of residuals B) Normality of residuals C) Independence of predictors D) Correlation among predictors January 2025 Marina NIKOLIDAKI Answer  B) Normality of residuals January 2025 Marina NIKOLIDAKI  The Durbin-Watson test assesses: A) Multicollinearity among predictors B) Autocorrelation in residuals Question C) Goodness-of-fit for linear regression models D) Variance of predictors January 2025 Marina NIKOLIDAKI Answer  B) Autocorrelation in residuals January 2025 Marina NIKOLIDAKI  Multicollinearity affects regression models by: Question A) Inflating standard errors of coefficients B) Increasing residual variance C) Improving model generalizability D) Enhancing predictor independence January 2025 Marina NIKOLIDAKI Answer  A) Inflating standard errors of coefficients January 2025 Marina NIKOLIDAKI  Question VIF values indicate multicollinearity when: A) VIF = 1 suggests correlation among predictors B) VIF > 5 suggests no correlation C) VIF > 10 suggests high multicollinearity D) VIF < 1 suggests perfect independence January 2025 Marina NIKOLIDAKI Answer  C) VIF > 10 suggests high multicollinearity January 2025 Marina NIKOLIDAKI  Residual plots are used to detect: A) Patterns indicating non-linearity or Question heteroscedasticity B) Independence among residuals C) Correlations between predictors D) Normality of residuals January 2025 Marina NIKOLIDAKI   A) Patterns indicating non-linearity or Answer heteroscedasticity January 2025 Marina NIKOLIDAKI  Outliers can be identified by: A) Large residual values in a residual plot Question B) High correlation coefficients in predictor analysis C) Low p-values in coefficient estimates D) High VIF values January 2025 Marina NIKOLIDAKI Answer  A) Large residual values in a residual plot January 2025 Marina NIKOLIDAKI  Question A regression model's R-squared value measures: A) The percentage of variance explained by the model B) The significance of coefficients C) The independence of predictors D) The normality of residuals January 2025 Marina NIKOLIDAKI   A) The percentage of variance explained by the Answer model January 2025 Marina NIKOLIDAKI