Wayground logo

Free Printable Worksheets

Font size

S
M
L
XL
Worksheets

Certified Legal Analytics Pretest

Total questions: 100

Worksheet time: 50mins

Name
Class
Date
1.

What is the primary purpose of legal analytics?

a)

Automating court rulings

b)

Enhancing legal decisions through data-driven insights

c)

Replacing lawyers entirely

d)

Standardizing all global legal systems

2.

Which is an example of legal analytics?

a)

Writing new laws

b)

Forecasting litigation outcomes with past case data

c)

Drafting contracts manually

d)

Verifying client identities

3.

Legal analytics primarily shifts legal practice from:

a)

Evidence-based to intuition-based

b)

Intuition-based to evidence-based

c)

Civil law to common law

d)

Statutory law to constitutional law

4.

Preprocessing of legal data involves:

a)

Drafting pleadings

b)

Cleaning, deduplicating, and standardizing datasets

c)

Writing statutes

d)

Creating legal arguments

5.

A key ethical challenge in legal analytics is:

a)

Overuse of case law

b)

Bias in datasets and predictive models

c)

Lawyers refusing to learn coding

d)

Too much client data digitization

6.

Classification in ML assigns:

a)

Continuous numerical predictions

b)

Items into predefined categories

c)

Similar items into clusters

d)

Features into reduced dimensions

7.

Predicting damages in a civil case uses:

a)

Classification

b)

Regression

c)

Clustering

d)

Association

8.

Grouping similar contracts without labels is:

a)

Classification

b)

Regression

c)

Clustering

d)

Dimensionality reduction

9.

The 'black box' problem refers to:

a)

Opaque decision-making in complex ML models

b)

Courts rejecting algorithms

c)

Lack of legal precedents

d)

Lawyers' poor coding skills

10.

Explainable AI (XAI) exists to:

a)

Make predictions faster

b)

Improve transparency and interpretability of ML models

c)

Remove need for lawyers in litigation

d)

Ensure bias in predictions

11.

Quantitative Legal Prediction (QLP) refers to:

a)

Automating legislation

b)

Empirical forecasting of legal outcomes

c)

Eliminating precedent reliance

d)

Teaching lawyers coding

12.

A known example of QLP is:

a)

Lex Machina's IP analytics platform

b)

Law school student ranking

c)

International law drafting

d)

Supreme Court appointments

13.

Research shows predictive algorithms:

a)

Often outperform experts in certain legal contexts

b)

Are less accurate than lawyers

c)

Match lawyers equally

d)

Cannot process case law

14.

One advantage of predictive analytics in law is:

a)

Removing client-lawyer interaction

b)

Faster and more cost-effective analysis

c)

Eliminating human strategy

d)

Creating global uniform law

15.

Adoption of QLP globally is slowed by:

a)

Lack of tech infrastructure and cultural acceptance

b)

Too many predictive tools

c)

Client disinterest in data

d)

Lack of case law

16.

High bias typically results in:

a)

Overfitting

b)

Underfitting

c)

Perfect predictions

d)

Balanced outcomes

17.

Precision measures:

a)

True positives over predicted positives

b)

True positives over actual positives

c)

False negatives over true negatives

d)

Cases filed over cases decided

18.

Recall measures:

a)

True positives over actual positives

b)

True negatives over false negatives

c)

Predicted positives over actual positives

d)

Accuracy of clustering

19.

Curse of dimensionality means:

a)

Too many features reduce model effectiveness

b)

Too few features

c)

Lack of training data

d)

Judicial overload

20.

Balancing bias and variance aims at:

a)

Optimal model performance

b)

Removing dimensionality

c)

Full accuracy

d)

Predicting without data

21.

Overfitting occurs when a model:

a)

Learns noise along with patterns

b)

Generalizes well to new data

c)

Ignores training data

d)

Is too simple

22.

Underfitting occurs when a model:

a)

Captures all noise

b)

Is too simple to capture data complexity

c)

Performs perfectly

d)

Has zero error

23.

Cross-validation is mainly used to:

a)

Prevent bias

b)

Assess generalizability of a model

c)

Automate predictions

d)

Improve case law analysis

24.

A model that performs well on training but poorly on test data is:

a)

Overfit

b)

Underfit

c)

Balanced

d)

Unbiased

25.

Cross-validation improves:

a)

Data bias

b)

Model stability and reliability

c)

Human interpretation

d)

Ethical review

26.

Logistic regression is best for:

a)

Predicting binary outcomes

b)

Predicting continuous values

c)

Grouping clusters

d)

Dimensionality reduction

27.

Maximum likelihood estimation finds parameters that:

a)

Minimize error

b)

Maximize probability of observing data

c)

Eliminate variance

d)

Automate law drafting

28.

In legal prediction, logistic regression could predict:

a)

Whether a judge rules in favor of plaintiff or defendant

b)

Trial duration

c)

Case filing dates

d)

Type of precedent applied

29.

A logistic regression output is:

a)

Probability between 0 and 1

b)

Continuous value

c)

Category clustering

d)

Raw likelihood

30.

Maximum likelihood is important because:

a)

It provides most probable parameter estimates

b)

It avoids model bias

c)

It eliminates overfitting

d)

It standardizes global laws

31.

KNN classifies data by:

a)

Decision trees

b)

Similarity to nearest neighbors

c)

Regression coefficients

d)

Random clustering

32.

Naive Bayes assumes:

a)

Features are dependent

b)

Features are independent

c)

Features are irrelevant

d)

Labels determine features

33.

A strength of KNN is:

a)

Simplicity and non-parametric nature

b)

Perfect interpretability

c)

Infinite scalability

d)

Zero data requirement

34.

A weakness of Naive Bayes is:

a)

Poor with correlated features

b)

High variance

c)

Overfitting

d)

High dimensionality always

35.

In contract classification, Naive Bayes would:

a)

Predict contract type using clause frequencies

b)

Predict damages

c)

Forecast trial length

d)

Clean data

36.

Decision trees work by:

a)

Splitting data into branches using features

b)

Using regression coefficients

c)

Random sampling

d)

Black-box deep models

37.

A key advantage of decision trees is:

a)

Interpretability

b)

Low variance

c)

Elimination of data preprocessing

d)

Perfect predictions

38.

Overfitting in decision trees is reduced by:

a)

Pruning

b)

Expanding

c)

Increasing depth

d)

Ignoring features

39.

Binary classification in trees involves:

a)

Two possible outcomes per case

b)

Unlimited outcomes

c)

Regression models

d)

Probabilistic clustering

40.

A weakness of decision trees is:

a)

Instability to small data changes

b)

Perfect stability

c)

Lack of visual clarity

d)

Complexity

41.

Ensemble methods combine:

a)

Multiple models for better performance

b)

Raw data directly

c)

Features without preprocessing

d)

Laws and statutes

42.

Random Forest reduces variance by:

a)

Averaging predictions from multiple trees

b)

Increasing depth of one tree

c)

Using regression instead

d)

Ignoring features

43.

Bagging in ML is:

a)

Bootstrap aggregation

b)

Bias elimination

c)

Regression method

d)

Dimensionality reduction

44.

Boosting differs by:

a)

Combining weak learners sequentially

b)

Randomizing features

c)

Only using one strong learner

d)

Removing variance

45.

Random Forests are widely used because:

a)

They balance accuracy and interpretability

b)

They eliminate data bias

c)

They require no computation

d)

They always outperform neural networks

46.

K-Means clustering works by:

a)

Assigning points to the nearest centroid

b)

Building decision trees

c)

Predicting binary outcomes

d)

Creating regression lines

47.

The main input needed for K-Means is:

a)

Number of clusters (k)

b)

Regression coefficient

c)

Decision boundaries

d)

Class labels

48.

Hierarchical clustering differs because:

a)

It creates a tree-like structure (dendrogram)

b)

It requires predefined k

c)

It cannot group data

d)

It is always supervised

49.

In legal analytics, clustering can help:

a)

Group similar cases or contracts

b)

Decide trial outcomes

c)

Replace lawyers

d)

Draft laws automatically

50.

A limitation of K-Means is:

a)

Sensitivity to initial centroid placement

b)

It cannot run on computers

c)

It only works on text

d)

It ignores numeric data

51.

Data visualization helps legal professionals:

a)

Identify patterns and communicate insights clearly

b)

Draft contracts

c)

File lawsuits faster

d)

Reduce statutes

52.

A popular R package for visualization is:

a)

ggplot2

b)

numpy

c)

pandas

d)

keras

53.

Good legal data visualization should be:

a)

Clear, accurate, and tailored to audience

b)

Decorative only

c)

Data-heavy with no interpretation

d)

Always in 3D

54.

Visualization in legal analytics may include:

a)

Case trends by year

b)

Lawyer's handwriting

c)

Number of judges per district

d)

Courtroom seating charts

55.

One advantage of visualization is:

a)

Makes complex data accessible and interpretable

b)

Increases data storage

c)

Guarantees perfect predictions

d)

Reduces need for data cleaning

56.

Preprocessing in R often uses:

a)

dplyr for filtering, selecting, and mutating data

b)

keras for clustering

c)

Excel only

d)

SQL exclusively

57.

Removing duplicates from legal data is:

a)

Preprocessing step

b)

Visualization step

c)

Regression step

d)

Prediction step

58.

Normalizing contract dates ensures:

a)

Standard format across datasets

b)

Better legal argument

c)

Judge's fairness

d)

Faster trials

59.

Tokenizing legal text is part of:

a)

Data cleaning

b)

Visualization

c)

Regression

d)

Overfitting

60.

A core benefit of preprocessing is:

a)

Increases accuracy of ML models

b)

Removes the need for models

c)

Eliminates case law

d)

Bypasses ethics concerns

61.

Network analysis studies:

a)

Relationships and connections between entities

b)

Regression coefficients

c)

Decision tree splits

d)

Random forests

62.

In law, network analysis can map:

a)

Judicial citation networks

b)

Criminal handwriting styles

c)

Lawyer salaries

d)

Number of trial rooms

63.

A node in network analysis could represent:

a)

A judge, case, or statute

b)

A regression line

c)

An error term

d)

A cluster centroid only

64.

An edge in legal networks represents:

a)

Relationship such as citations or references

b)

Outcome of case

c)

Model bias

d)

Lawyer's income

65.

Visualizing legal networks helps:

a)

Detect influential cases/judges

b)

Automate contract drafting

c)

Eliminate all precedents

d)

Remove need for lawyers

66.

NLP in law is mainly used for:

a)

Analyzing legal text at scale

b)

Predicting judge salaries

c)

Drafting new statutes

d)

Filing cases automatically

67.

Tokenization in NLP is:

a)

Splitting text into words or phrases

b)

Building regression models

c)

Clustering outcomes

d)

Cleaning data duplicates

68.

Sentiment analysis of court opinions can:

a)

Detect tone or bias in judgments

b)

Replace judicial reasoning

c)

Predict damages directly

d)

Remove case citations

69.

Named Entity Recognition (NER) in law identifies:

a)

Parties, dates, clauses in documents

b)

Regression outputs

c)

Judge's accuracy

d)

Random clusters

70.

A limitation of NLP in law is:

a)

Ambiguity and complexity of legal language

b)

Data always structured

c)

No text data in law

d)

Judges rejecting text

71.

Contract analytics tools can:

a)

Flag non-standard clauses automatically

b)

Write laws

c)

Predict trial duration

d)

Eliminate law firms

72.

SEC data analysis in law often focuses on:

a)

Corporate filings & compliance monitoring

b)

Civil case outcomes

c)

Courtroom seating

d)

Drafting statutes

73.

Judicial prediction tools analyze:

a)

Judges' past rulings and behavior

b)

Lawyers' handwriting

c)

Court architecture

d)

Jury size

74.

Regulatory analytics helps:

a)

Anticipate compliance risks

b)

Draft contracts manually

c)

Predict personal income

d)

Avoid visualizations

75.

Sentiment analysis in legal analytics can:

a)

Gauge public or judicial attitude from texts

b)

Predict numeric damages

c)

Remove precedents

d)

Automate law drafting

76.

Contract review platforms like Luminance use:

a)

Machine Learning + NLP

b)

Only human reading

c)

Only Excel

d)

Only clustering

77.

SEC filing analysis supports:

a)

Investor protection and fraud detection

b)

Criminal handwriting analysis

c)

Drafting constitutions

d)

Jury training

78.

Predictive justice platforms can:

a)

Forecast outcomes with probability scores

b)

Guarantee wins in all cases

c)

Replace courts entirely

d)

Eliminate appeals

79.

A company can use litigation trend analysis to:

a)

Anticipate rising claims and adjust policies

b)

Increase staff salaries

c)

Build new courthouses

d)

Replace HR

80.

Applied analytics in law firms helps:

a)

Resource allocation and client advisory

b)

Erase statutes

c)

Remove precedent law

d)

Automate judgments

81.

SVM works by:

a)

Finding the optimal hyperplane separating classes

b)

Clustering similar items

c)

Regression only

d)

Decision trees

82.

A strength of SVM is:

a)

Effective in high-dimensional spaces

b)

Perfect interpretability

c)

No preprocessing required

d)

Always unsupervised

83.

Kernel trick in SVM allows:

a)

Non-linear separation by mapping data to higher dimensions

b)

Eliminating features

c)

Contract drafting

d)

Removing cases

84.

A limitation of SVM is:

a)

Computationally intensive on large datasets

b)

Always inaccurate

c)

Cannot classify

d)

Only works on text

85.

In legal analytics, SVM could classify:

a)

Case outcomes based on multiple features

b)

Judge's handwriting

c)

Trial dates

d)

Courtroom locations

86.

EM algorithm is used for:

a)

Finding parameters in models with hidden variables

b)

Clustering judges manually

c)

Visualizing contracts

d)

Cleaning text

87.

EM is often applied in:

a)

Mixture models like Gaussian Mixtures

b)

Linear regression

c)

Decision trees

d)

Random forests

88.

EM works through:

a)

Alternating expectation and maximization steps

b)

Only regression steps

c)

One-time parameter estimation

d)

Tokenization

89.

In law, EM can help with:

a)

Modeling hidden patterns in contract data

b)

Predicting trial locations

c)

Writing statutes

d)

Clustering lawyers manually

90.

A drawback of EM is:

a)

Can converge to local optima

b)

Always finds global optimum

c)

No iterative process

d)

Eliminates datasets

91.

Neural networks are inspired by:

a)

Biological neurons in the brain

b)

Court hierarchies

c)

Legal statutes

d)

Random forests

92.

A perceptron is:

a)

Basic unit of a neural network

b)

A clustering centroid

c)

A regression line

d)

A law article

93.

Deep learning refers to:

a)

Neural networks with many hidden layers

b)

Only decision trees

c)

Regression only

d)

Random clustering

94.

In law, neural networks can be applied to:

a)

Analyzing large volumes of case text

b)

Writing laws

c)

Drafting judgments directly

d)

Filing lawsuits

95.

A weakness of neural networks is:

a)

Interpretability ('black box' problem)

b)

Perfect clarity

c)

No training needed

d)

Always linear

96.

MLaaS stands for:

a)

Machine Learning as a Service

b)

Machine Learning and Algorithmic Statistics

c)

Machine Learning Against Systems

d)

Machine Law at Scale

97.

A benefit of MLaaS is:

a)

Scalability and reduced infrastructure costs

b)

No need for lawyers

c)

No ethical concerns

d)

Always interpretable

98.

Cloud-based ML platforms support:

a)

On-demand training and deployment of models

b)

Only Excel-based models

c)

Manual research

d)

Handwritten judgments

99.

Future trends in legal tech include:

a)

Integration of AI with legal workflows

b)

Elimination of law schools

c)

Manual-only research

d)

Removal of precedents

100.

The shifting economics of legal analytics implies:

a)

Data-driven firms gain competitive advantage

b)

All firms must charge higher fees

c)

Clients ignore predictions

d)

Technology has no role