F3 CHIA Guide Informatics Theory Flashcards
(342 cards)
1
Q
What is a concept?
A
An idea that has been formally developed and organised
Concepts form the building blocks of theories.
2
Q
What is a theory?
A
A set of concepts, models, principles, definitions that make sense of a phenomenon by determining relationships among variables
Theories are established and validated by experiments and evidence.
3
Q
What is the purpose of a conceptual framework?
A
It serves as a roadmap for your study, helping you visualise your research project and put it into action
It defines relevant variables and maps out relationships.
4
Q
How is a conceptual framework used in quantitative studies?
A
To determine survey questions or data points, or to generate a hypothesis for explanations and predictions.
5
Q
How is a conceptual framework used in qualitative studies?
A
To provide a working hypothesis or a set of research questions, or to identify or explore categories in descriptive research.
6
Q
What is the purpose of a theoretical framework?
A
To introduce and describe the theory/theories underpinning the research problem.
7
Q
How many theoretical frameworks might be involved in master’s research?
A
One or two theoretical frameworks.
8
Q
How many theoretical frameworks might be involved in PhD research?
A
Three or more theoretical frameworks.
9
Q
What is a key difference between a conceptual framework and a theoretical framework?
A
A conceptual framework is more about the approach taken in answering a research question, while a theoretical framework is developed from existing theories.
10
Q
What is the composition of a conceptual framework?
A
It is composed of several concepts and may include a theoretical framework.
11
Q
What does a theoretical framework derive from?
A
It is derived from theory.
12
Q
What factors can a conceptual framework identify?
A
Factors influencing a particular field, such as exploration of ‘masquerade’ mimicry in animals.
This can include phenomena like protective mimicry, crypsis, and aposematism.
13
Q
What is an example of a theoretical framework?
A
Darwin’s theory of evolution by natural selection.
14
Q
When should a conceptual framework be created?
A
Before starting experiments.
15
Q
What format can a conceptual framework take?
A
It can be in a written or diagrammatic format.
16
Q
What is the first step in constructing a theoretical framework?
A
Read and review the literature to identify long-standing themes and the main concerns for the inquiry.
17
Q
Fill in the blank: A conceptual framework may include a _______.
A
[theoretical framework]
18
Q
True or False: A theoretical framework can be based on outcomes from a single study.
A
False
19
Q
What does a conceptual framework help you to do?
A
Visualise linkages between various concepts and theories.
20
Q
What can a framework make research findings more?
A
Meaningful.
21
Q
What is a model in scientific contexts?
A
A representation of reality
Models can also be considered abstractions of reality.
22
Q
What are the two major types of models discussed?
A
- Static models
- Dynamic models
23
Q
Define a static model.
A
A model without state changes that always stays in one state.
24
Q
Define a dynamic model.
A
A model where the state changes over time.
25
What is a physical model?
A model that exists in the physical world and can be observed.
26
What is a mental model?
An abstract model composed of thoughts and ideas.
27
What is a model description?
An artifact that describes a model.
28
True or False: A model description is a model itself.
False
29
What concept is used to relate models to reality?
The concept of a system.
30
Define a system.
A potentially changing set of objects and their properties that interact with each other.
31
What is perspective in the context of this study?
A filter applied to reality by our perception and a structure imposed on reality using our concepts.
32
What is the difference between external and social worlds?
The external world is reality, while the social world is created by humans.
33
What are the three types of realities discussed?
* Physical realities
* Digital realities
* Mental realities
34
What is an example of a physical system?
A wolf-sheep ecosystem.
35
What is a digital system?
A system where objects are digital entities.
36
What is a mental system?
A system where the contained objects are thoughts.
37
Fill in the blank: A __________ model refers to models composed of thoughts and ideas.
mental
38
True or False: Digital systems can be controlled in ways that physical systems cannot.
True
39
What is an important aspect of semiotics in this study?
It provides a philosophical underpinning for understanding models and their applications.
40
What is the significance of interdisciplinarity mentioned in the study?
It allows for the diffusion of knowledge across different scientific fields.
41
What is the focus of the field of semiotics?
Exploring everything that makes meaning and signals for something else.
42
What are examples of physical models?
* Scale models
* Dolls
* Computer simulations
43
What is object-oriented in the context of perception of reality?
Phenomena and concepts are represented by objects and classes.
44
What limits our perception of reality?
Senses, imagination, concepts, and purpose.
45
What does the term 'perspective' encompass?
Sense limitations, concepts, and purpose.
46
What does the study aim to unify?
Models across different disciplines.
47
What role does communication play in the context of models?
It is essential for expressing and understanding models and systems.
48
What are the contained objects in a mental system?
Thoughts
## Footnote
A mental system can simulate and run experiments, leading to outcomes like peaks of sheep every 6 years.
49
How can an experienced programmer predict the outcome of a digital wolf-sheep system?
By running it in her brain
## Footnote
This creates a mental wolf-sheep system that allows for experimentation.
50
What does a mental system represent?
A simulation or imagination of reality in the mind of a person
## Footnote
It evolves over time and allows control over its evolution.
51
What determines the perspectives available in a mental system?
The concepts available in the mind of the person
## Footnote
Different concepts can lead to different mental systems triggered by the same object.
52
What is the difference between a static system and a snapshot?
A static system is a long sequence of the same snapshots
## Footnote
A snapshot shows the structure of the system at a specific point in time.
53
What is an example of a static system?
The wolf-sheep system when no changes occur
## Footnote
If wolves catch all sheep, the system becomes static.
54
How can a perspective influence the classification of systems?
It can make a difference between static and dynamic systems
## Footnote
This is based on whether changing elements of reality are considered.
55
What is a model?
A system that is in the model-of relationship to a referent system
## Footnote
The model is analogous to the referent system.
56
What is the relationship between the digital, mental, and physical wolf-sheep systems?
They capture the same essence and are models of each other
## Footnote
The digital system is a model of the mental system, which is a model of the physical system.
57
What defines a system description?
A set of statements about a system, given in some language
## Footnote
It should be complete and well-formed to describe the system's structure and dynamics.
58
What are different forms of system descriptions?
Text, pictures, diagrams, formulas, icons
## Footnote
These representations can imply one or more systems.
59
True or False: Static systems do not change at all.
False
## Footnote
Even static systems can have underlying dynamic processes that are not visible from a specific perspective.
60
Fill in the blank: A _______ is a system that has a perspective based on its purpose.
[model]
## Footnote
The purpose of the model restricts the referent system.
61
How can models and systems be considered similar?
They allow transitivity for the model-of relationship
## Footnote
This is discussed in relation to the definitions in the literature.
62
What is the significance of static systems in science?
They capture ground truths about the world
## Footnote
Formulas in physics and chemistry are examples of static systems.
63
What is the role of perspective in understanding static and dynamic systems?
It determines whether the dynamicity of the system is relevant
## Footnote
Different perspectives can reveal or obscure changes in the system.
64
What is an example of a mental model?
A taxonomy of concepts
## Footnote
This can also be depicted in formal systems like UML diagrams.
65
What is the main difference between systems and descriptions?
Descriptions describe (or create or refer to) systems.
66
True or False: Two descriptions can describe the same system.
True
67
What does the term 'description' denote in this context?
A set of statements about the system or a semiotic representation referring to it.
68
Fill in the blank: The distinction between a description and the real item is famously highlighted in the painting '________.'
The Treachery of Images
69
What is an example of a descriptive description?
A sketch or 3D model of a building.
70
What are the two types of model descriptions?
* Descriptive
* Prescriptive
71
What is the meaning defined on the language level (L) related to?
It defines what descriptions mean.
72
What is the difference between design time and run time?
* Design time: when the description is created
* Run time: when the description is used to create a system
73
True or False: Descriptions and systems are often disregarded as different entities.
True
74
What does Peirce's semiotics focus on?
The relations between the world, the sign, and the human users of the sign.
75
What are the two components of a sign according to semiotics?
* Expression (what we can sense)
* Content (what the sign refers to)
76
Fill in the blank: Semiotics looks at the relations between the world, the sign, and the ________.
human users of the sign
77
What is a model-of relationship?
A relationship where two systems are analogous in some way.
78
What is meant by 'static' and 'dynamic' aspects in systems?
* Static: artifacts that do not change
* Dynamic: systems that evolve over time
79
What does Shannon and Weaver's model of communication emphasize?
The transmission of a message from a sender to a receiver.
80
What is the difference between semantics and meaning in communication?
* Semantics: formal system of signification
* Meaning: individual understanding in context
81
What is the role of communication in semiotics?
It is crucial for creating the concepts that shape the mental system.
82
What does a representation stand for in semiotics?
Something else.
83
According to Peirce, what is the interpretant?
The human interpretation of the sign.
84
Fill in the blank: A sign is an abstract entity that can be ________ in many ways.
realized
85
What is the connection between the sign and the mediate object in Peirce's semiotics?
The sign refers to the immediate object experienced by humans.
86
What are the two types of objects Peirce distinguishes in his semiology?
* Mediate object
* Immediate object
87
What is the focus of this study regarding communication?
The result of the meaning-making process.
88
What is the significance of a diagram in describing a dynamic system?
It can represent a complete run and trace system states over time.
89
What is the focus of the study regarding meaning-making?
The study focuses on the part of the process where a person forms a mental system based on a verbal or non-verbal description of a system.
90
How does perspective relate to mental models?
Perspective provides the context of a system, shaping what is possible to observe in reality.
91
What does FRISCO identify models as?
FRISCO identifies models to be conceptions.
92
What distinguishes a physical system from a mental model?
A physical system is a passive interpretation of reality, while a mental model is an active mental system that can predict outcomes.
93
Fill in the blank: A _______ is a mental system that can help us handle reality by predicting events.
mental model
94
What are UML models described as in the context of modeling languages?
UML models are collections of UML diagrams.
95
What is validation in the context of systems?
Validation is the process of finding out whether a system has the right model-of relation to an existing or planned real system.
96
How are descriptions and representations differentiated?
A description is a set of statements about something, while a representation stands for something else.
97
True or False: Descriptions are generally understood as dynamic.
False
98
What does the term 'dynamic description' refer to?
Dynamic descriptions refer to descriptions that change, such as animated charts.
99
What is a key element in distinguishing between static and dynamic systems?
A dynamic system has several different system states over time, while a static system has the same system state at each time point.
100
Fill in the blank: In the context of FRISCO, both systems and models are considered _______.
models
101
What influences our perspective of the world according to the study?
Our understanding of the world influences our perspective.
102
What do object diagrams represent in UML?
Object diagrams represent instance specifications of systems.
103
How does the study define a model?
A model is a system that is analogous to another system, called a referent system.
104
What is the relationship between concepts and mental models?
Concepts are tied to our experiences and used to build our mental models.
105
Fill in the blank: Each linguistic term is a _______ in the brain by means of its expression.
concept
106
What does the term 'meta-description' refer to?
A meta-description is a description of a description.
107
What is a challenge in using the term 'representation' in the context of this study?
The term 'representation' can imply a more formal and complete understanding than 'description'.
108
In media studies, which media are considered dynamic?
Radio, TV, and film are considered dynamic media.
109
What is the significance of the model-of relationship?
The model-of relationship connects models to their referent systems.
110
True or False: Static vs. dynamic systems are exclusive categories.
False
111
What does the study suggest about the nature of descriptions?
Descriptions can be static or dynamic depending on the perspective.
112
What is the difference between reality and systems?
Systems are parts of reality observed using a perspective.
113
What distinguishes dynamic systems from static systems?
Dynamic systems have several different states over time, while static systems always have the same state at each time point.
114
Define a model in the context of systems.
A model is a system that is analogous to another system, called a referent system.
115
What is the significance of perspective in observing systems?
The chosen perspective can lead to observing systems that do or do not change.
116
What is a description of a system?
A collection of statements about a system, shaped by the language it is presented in.
117
True or False: Descriptions of systems are the same as the systems themselves.
False.
118
How do descriptions relate to systems?
Descriptions lead to systems by their implicit or explicit meaning derived from the language used.
119
What does the term 'model-of relationship' refer to?
The comparison of an implied system to another system.
120
Fill in the blank: A _______ system always has the same system state at each time point.
static
121
Fill in the blank: A _______ system has several different system states over time.
dynamic
122
What is the purpose of descriptions in the context of systems?
To communicate about systems.
123
What do the descriptions of models represent?
They are not the models themselves but the descriptions of the models.
124
What is the importance of using a matching perspective for systems?
It brings about similarity between the model and referent system.
125
What is meant by 'snapshot' in the context of static systems?
A snapshot refers to the constant state of a static system at a given time.
126
What type of systems can be observed depending on the perspective?
Both dynamic and static systems.
127
What role does language play in the description of systems?
It shapes the collection of statements that describe the system.
128
True or False: Models can be used in diverse domains.
True.
129
Who are the authors of the document?
Henderik Proper and András J. Molnár
130
What is the affiliation of Henderik Proper?
Vienna University of Technology, Austria
131
What is the affiliation of András J. Molnár?
Computer and Automation Research Institute (MTA), Hungary
132
What section comes after 'Models' in the table of contents?
'Descriptions'
133
What percentage of researchers rate the articles as excellent or good?
94%
134
True or False: The document includes a section on 'Discussion'.
True
135
What is the purpose of the research integrity team mentioned in the document?
To safeguard the quality of each article we publish.
136
Fill in the blank: The document includes a section on _______.
[Communication and semiotics]
137
What is one of the listed articles that people also looked at?
The effects of social density, spatial density, noise, and office views on perceived personal space in the virtual workplace
138
What type of modeling is discussed in the article by Azeem Hafeez and others?
Distortion modeling and neural networks
139
What is one of the guidelines mentioned in the document?
Author guidelines
140
What section follows 'Conclusions' in the table of contents?
'Data availability statement'
141
True or False: The document has a section titled 'Conflict of interest'.
True
142
What is one of the formats available for exporting citations?
BibTex
143
What publication is associated with the authors?
Frontiers Media S.A.
144
Fill in the blank: The document mentions a section on _______ ethics.
[Policies and publication ethics]
145
What is the primary focus of the article by Sofia Kohan and others?
Model-centered approaches to conceptual modeling of IoT systems
146
What is SOA?
Service-oriented architecture (SOA) is a software design approach that focuses on building functional, scalable software systems from individual components called services.
147
What are the key emphases of SOA?
The key emphases of SOA are modularity, reusability, and interoperability.
148
How do services interact in SOA?
Services can interact with one another to perform tasks, such as allowing single sign-on access to various business applications.
149
What is the role of a registry in SOA?
Services are published to a registry, and applications request the latest version of a service from this registry.
150
What are the benefits of using SOA?
* Faster development cycles
* Easier maintenance
* Adaptability
* Scalability
151
What does loosely coupled mean in the context of SOA?
Individual services are loosely coupled and can communicate and transmit data as needed without being tightly integrated.
152
What is an example of SOA in cloud software development?
An example is creating applications with universal logins, where authentication management is handled by a set of services.
153
What is an example of a service in SOA?
Examples include payment processing services, customer management services, and product recommendation services.
154
What is a key benefit of switching services in SOA?
Switching services, such as payment processors, can be done without altering the applications that use the service.
155
What are the main characteristics of a service-oriented architecture?
* Loose coupling of independent services
* Interoperability
* Scalability
* Offloaded maintenance
156
How does SOA facilitate easier maintenance?
SOA allows individual services to be independently maintained without impacting the larger project.
157
Fill in the blank: SOA preserves _____ compatibility and facilitates future planning.
backward
158
What differentiates SOA from microservices?
SOA considers use on a higher, expansive level, while microservices focus on application-level tasks and provide specialized functions.
159
Where can SOA be used?
SOA can be used for nearly any application as long as standalone services exist to fulfill the system's requirements.
160
True or False: SOA requires developers to maintain the services they use.
False
161
What is an example of a consumer application using SOA?
An app for runners that integrates GPS and map services.
162
What is the modularity characteristic of SOA?
SOA allows services to be developed, tested, and deployed independently.
163
What is the primary goal of clinical prediction models?
To forecast future health outcomes given a set of baseline predictors to facilitate medical decision making and improve health outcomes.
164
How many prediction models were identified in obstetrics according to a review?
263 prediction models.
165
What has heightened interest in predicting health outcomes?
The increasing availability of big data.
166
What does the PROGRESS framework provide?
Detailed guidance on different types of prognostic research.
167
What does the TRIPOD statement recommend?
Recommendations for reporting prediction models, including those in clustered datasets.
168
What is the purpose of the PROBAST tool?
To provide a structured way to assess the risk of bias in a prediction modeling study.
169
What is a common methodological limitation in published prediction modeling studies?
Severe methodological shortcomings that undermine their usefulness.
170
What is overfitting in the context of prediction modeling?
Estimating many model parameters from few data points, leading to overestimating the model’s performance.
171
What are the first steps in developing a clinical prediction model?
Define aims, create a team, review literature, and start writing a protocol.
172
What should be clearly defined when starting a prediction model?
The target population, the outcome to be predicted, the healthcare setting, intended users, and clinical decisions the model will inform.
173
What is the training set in prediction modeling?
Data used to develop a model.
174
What does 'discrimination' refer to in prediction models?
The capacity of the model to rank patients concerning their outcomes.
175
What does calibration assess in a prediction model?
The agreement between predicted and observed outcomes.
176
What is underfitting?
When a model is not complex enough to capture patterns in the data well.
177
What is apparent performance?
Performance of a prediction model when the same dataset is used for developing and assessing performance.
178
What does the bias-variance trade-off relate to?
The trade-off between having a simple, underfitting prediction model versus a complex, overfitting one.
179
What is internal validation?
Methods for obtaining an honest assessment of the performance of prediction models using the data it was developed with.
180
What is external validation?
Evaluation of model’s performance in new data not used for training the model.
181
What is penalisation in prediction modeling?
A method for reducing model complexity to obtain better predictions.
182
What are LASSO, ridge regression, and elastic net?
Penalised estimation methods for regression models.
183
What does reproducibility mean in the context of prediction models?
Estimated model performance can be reproduced in a new sample from the same population or setting.
184
Fill in the blank: The _______ refers to the ability of the model to produce accurate predictions in new patients drawn from a different but related population or setting.
Transportability
185
True or False: The first step in developing a prediction model is to assemble a team with expertise in the specific medical field.
True
186
What is a common pitfall in prediction modeling?
Inappropriate categorizing of continuous outcomes or predictors.
187
What is the focus of the article discussed?
Methods for predicting a future health outcome.
188
What is the purpose of a scoping review in the context of developing prediction models?
To identify relevant published prediction models and studies on important risk factors
189
What is external validation in prediction modeling?
Assessing the validity of an existing model for the intended setting
190
Name common strategies for updating a prediction model.
* Recalibration
* Revision
* Extension
191
What is the advantage of using time-to-event data over binary outcomes in prediction models?
Time-to-event data provides richer information, such as survival probability at any time point
192
What should candidate predictors ideally be based on?
Literature review and expert knowledge
193
Why should we avoid dichotomizing or categorizing continuous predictors?
It reduces information and diminishes statistical power
194
What is the significance of considering the model's intended use?
It influences the availability of data and the selection of variables
195
What is a major drawback of using data from randomized clinical trials for prediction models?
Limited generalizability due to trial participants not representing the broader patient population
196
What is underfitting in the context of prediction models?
A very simple model or a model based on covariates that are not associated with the outcome
197
What is overfitting?
A model with too many predictors developed in a small dataset that performs well only in that dataset
198
What method is usually recommended for handling missing data during model development?
Multiple imputation
199
What is the purpose of the imputation model in multiple imputation?
To create several versions of the dataset with missing values imputed
200
What should the imputation model include?
The same predictors included in the final model used for making predictions
201
True or False: Single imputation is generally more consistent and stable than multiple imputation.
False
202
What is a potential issue with imputation methods?
They might fail when the tendency of the outcome to be missing correlates with the outcome itself
203
What is the TRIPOD statement used for?
Guidance when writing the study protocol for prediction models
204
Fill in the blank: A study protocol should guide subsequent steps and can be made publicly available in an _______.
open access journal
205
What is the main advantage of using registry data for developing prediction models?
Relatively large sample size and representativeness
206
What should be done if predictors have limited variation?
Exclude them from the model
207
What is the recommended approach when dealing with missing data during model use?
Impute data during the development and implementation phases
208
What is the focus of step 3 in developing prediction models?
Define the outcome measure
209
What does the PROBAST tool assist with?
Evaluating the risk of bias in prediction models
210
What is single imputation used for?
Single imputation can be used during model development and model use
## Footnote
It is a method to deal with missing data.
211
What might cause imputation methods to fail?
When the tendency of the outcome to be missing correlates with the outcome itself
## Footnote
Example: Patients receiving a new treatment might miss follow-up visits if the treatment was successful.
212
What are the usual starting points in modeling strategies for prediction?
* Linear regression for continuous outcomes
* Logistic regression for binary outcomes
* Cox or simple parametric models for survival outcomes
213
What is a competing event in prediction modeling?
A situation when several possible outcomes exist, but a person can only experience one event
## Footnote
Example: Death from another cause when predicting death from breast cancer.
214
Why are univariable selection methods not recommended?
They do not consider the association between predictors and could lead to loss of valuable information.
215
What is penalization in model estimation?
Adding penalty terms to the model to control the complexity and prevent overfitting.
216
What are some examples of penalization methods?
* Ridge
* LASSO (least absolute shrinkage and selection operator)
* Elastic net
217
What does LASSO and elastic net do in model estimation?
They can be used for variable selection by excluding some predictors by setting their coefficients to zero.
218
What is the purpose of combining multiple imputed datasets?
To develop different models for each modelling strategy and combine their predictions.
219
What are the two dimensions of prediction performance?
* Discrimination
* Calibration
220
How is discrimination assessed for continuous outcomes?
By contrasting predicted and observed outcomes directly.
221
What is the Brier score used for?
To measure overall performance by calculating the mean squared difference between predicted probabilities and actual outcomes.
222
What is internal validation?
A procedure that focuses on reproducibility and aims to ensure assessments of model performance are honest.
223
What is k-fold cross validation?
An approach where the data is divided into k subsets, and the model is built using k−1 of these subsets.
224
What is the risk of using the same dataset for model validation?
It may lead to overestimation of model performance.
225
What does transportability refer to in model validation?
The ability to produce accurate predictions in new patients drawn from a different but related population.
226
What is temporal validation in model performance assessment?
A method that informs about possible time trends in model performance but is not recommended for the development phase due to data splitting being arbitrary.
227
What is k-fold cross validation?
An approach where data is divided into k subsets; the model is built using k−1 folds and evaluated on the remaining fold, repeated for all folds.
228
What is bootstrapping in model validation?
A method that calculates optimism and optimism-corrected performance measures for any model, generally leading to more stable and less biased results.
229
What does model instability indicate?
It means small changes in the development dataset lead to large changes in model structure and predictions.
230
How is model stability assessed during development?
By using a bootstrap approach to create numerous models from bootstrap samples and comparing their predictions with the original model.
231
What are the steps for calculating optimism corrected measures of performance through bootstrapping?
1. Calculate apparent performance in the original sample.
2. Create bootstrap samples and construct models.
3. Calculate performance in the original sample.
4. Average optimism values.
5. Calculate corrected performance.
232
What is internal-external validation?
A method that partitions data into clusters and uses one cluster as the test set while training the model on remaining clusters, providing insights into model generalisation.
233
What is external validation?
Testing the model on a new set of patients not used for model development to determine its transportability before clinical implementation.
234
What does calibration drift refer to?
When a model's calibration is suboptimal across new settings or deteriorates over time.
235
What is the purpose of decision curve analysis?
To assess whether a prediction model should be used in practice by quantifying its clinical impact, considering benefits, risks, and costs.
236
Fill in the blank: A prediction model might be well calibrated, but its value depends on how we intend to use it in _______.
[clinical practice]
237
What does net benefit in decision analysis represent?
The expected percentage of true positives minus the expected percentage of true negatives, multiplied by a weight determined by the chosen cut-off threshold.
238
What are pitfalls in the interpretation of decision curves?
They cannot determine at what threshold probability the model should be used and can be affected by optimism.
239
What are some methods to assess the importance of individual predictors?
1. Estimated coefficients in linear regression models.
2. Fitting the model with and without the predictor.
3. Permutation importance algorithm.
4. SHAP (Shapley additive explanations).
240
What is the purpose of the TRIPOD reporting guideline?
To ensure all important aspects of the model development process and results are covered in the publication.
241
What is relapsing-remitting multiple sclerosis (RRMS)?
A chronic inflammatory disorder of the central nervous system characterized by attacks of worsening neurological function followed by periods of recovery.
242
What is the aim of the prediction model for RRMS described in the document?
To predict relapse within two years in patients with RRMS to inform treatment decisions.
243
What is the aim of the prediction model for RMMS?
To predict relapse within two years in patients with RRMS
## Footnote
This prediction can help in making informed treatment decisions.
244
What does a high risk of relapsing in RRMS patients imply for treatment?
Patients might consider intensifying treatment, such as using more active disease-modifying drugs or considering stem cell transplantation.
245
What kind of team was formed for the model development?
A multidisciplinary team comprising clinicians, patients, epidemiologists, and statisticians.
246
What were the limitations of existing prediction models identified in the literature review?
* Lack of internal validation
* Inadequate handling of missing data
* Lack of assessment of clinical utility
247
What outcome measure was chosen for the prediction model?
The occurrence of at least one relapse within a two-year period for people with RRMS.
248
List some predictors used in the model based on the literature review.
* Age
* Expanded disability status scale score
* Previous treatment for multiple sclerosis
* Months since last relapse
* Sex
* Disease duration
* Number of previous relapses
* Number of gadolinium enhanced lesions
249
What was the sample size of the data used in developing the model?
1752 observations from 935 patients.
250
What statistical model was developed for prediction?
A Bayesian logistic mixed effects prediction model.
251
What was the optimism corrected AUC value for the model?
0.65.
252
What does a decision curve analysis indicate about the model?
Deciding to intensify treatment based on the model is preferable to simpler strategies for thresholds between 15% and 30%.
253
Which factors were associated with higher odds of experiencing a relapse?
* Younger age
* Higher expanded disability status scale scores
* Shorter durations since the last relapse
254
What is the purpose of the R-shiny web application developed?
To allow patients, doctors, and decision-makers to estimate the probability of experiencing at least one relapse within the next two years.
255
What are the proposed steps for developing a clinical prediction model?
* Define aims
* Choose between developing a new or updating an existing model
* Define outcome measure
* List candidate predictors
* Collect data
* Consider sample size
* Deal with missing data
* Fit prediction model(s)
* Assess model performance
* Decide on the final model
* Perform a decision curve analysis
* Assess predictive ability of individual predictors
* Write up and publish
256
What does TRIPOD stand for in the context of prediction models?
Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis.
257
True or False: The tutorial provided is a complete and exhaustive guide to prediction models.
False.
258
Fill in the blank: The model's calibration slope was _______.
0.91.
259
What is the significance of bootstrapping in model development?
It is recommended as the method of choice for internal validation.
260
What is a potential issue raised regarding sample size calculations?
The minimum sample size indicated was larger than the available sample, raising concerns about possible overfitting.
261
What does the acronym PROGRESS stand for?
Prognosis Research Strategy
## Footnote
PROGRESS is a framework for researching clinical outcomes.
262
What is the TRIPOD statement focused on?
Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis
## Footnote
TRIPOD aims to improve the reporting quality of clinical prediction models.
263
What is the purpose of the PROBAST tool?
To assess the risk of bias and applicability of prediction model studies
## Footnote
PROBAST helps evaluate the methodological quality of prediction models.
264
What does the STRATOS initiative provide guidance on?
Validation of prediction models in the presence of competing risks
## Footnote
STRATOS focuses on methodological approaches for handling competing risks in predictive modeling.
265
What are the key components of external validation in prediction models?
* Model performance assessment
* Generalizability
* Applicability
## Footnote
External validation is crucial for determining how well a model performs on new data.
266
True or False: Dichotomizing continuous variables is generally a good practice in statistical analysis.
False
## Footnote
Dichotomizing can lead to loss of information and statistical power.
267
Fill in the blank: The _____ statement provides updated guidance for reporting clinical prediction models that use regression or machine learning methods.
TRIPOD+AI
## Footnote
The TRIPOD+AI statement addresses the integration of artificial intelligence in clinical prediction modeling.
268
What is the focus of the systematic review conducted by Damen et al. in 2016?
Prediction models for cardiovascular disease risk in the general population
## Footnote
This review summarizes existing prediction models and their effectiveness.
269
What is the main goal of the Predictive Approaches to Treatment effect Heterogeneity (PATH) statement?
To provide a framework for understanding treatment effect variability
## Footnote
PATH emphasizes the need for personalized medicine approaches.
270
What are some common errors in decision curve analysis?
* Misinterpretation of net benefit
* Ignoring model uncertainty
* Overlooking the importance of clinical context
## Footnote
Statistical thinking is crucial for accurate decision-making in prediction modeling.
271
What does the acronym BMJ stand for?
British Medical Journal
## Footnote
BMJ publishes research and reviews in the field of medicine.
272
What is the significance of the systematic reviews of prediction models mentioned by Collins et al.?
To synthesize evidence on the performance and applicability of clinical prediction models
## Footnote
Systematic reviews help in understanding the effectiveness of different predictive approaches.
273
What is the impact of sample size on predictive logistic regression models?
It affects model validity and generalizability
## Footnote
Adequate sample size is crucial for developing robust predictive models.
274
True or False: Machine learning methods require less data than traditional statistical methods.
False
## Footnote
Machine learning techniques often require large datasets to perform effectively.
275
What is the primary concern of the study by Ramspek et al. regarding external validation?
Understanding the methodology and reporting of external validation studies
## Footnote
External validation is essential for confirming the applicability of prediction models.
276
Fill in the blank: The _____ method is a statistical approach used for updating clinical prediction models.
Bayesian
## Footnote
Bayesian methods allow for the incorporation of prior information in model updating.
277
What is a common limitation of many clinical prediction models?
Overfitting to the training data
## Footnote
Overfitting can result in poor performance on new, unseen data.
278
What is the goal of the systematic review conducted by Meehan et al. in 2022?
To assess the progress and challenges of clinical prediction models in psychiatry
## Footnote
This review highlights advancements and areas needing improvement in psychiatric prediction modeling.
279
What is the main focus of the study by Harrell F. titled 'Regression Modeling Strategies'?
Applications to Linear Models, Logistic and Ordinal Regression, and Survival Analysis
## Footnote
This work is essential for understanding regression modeling techniques.
280
What is multiple imputation used for in epidemiological and clinical research?
Dealing with missing data
## Footnote
It addresses the potential and pitfalls associated with missing data.
281
Who conducted a systematic review on the risk of bias in studies using supervised machine learning techniques?
Andaur Navarro CL, Damen JAA, Takada T, et al.
## Footnote
This review highlights concerns regarding prediction model development.
282
What are competing risks in biostatistics?
Situations where different outcomes prevent the occurrence of the primary event of interest
## Footnote
Competing risks are crucial for understanding survival data.
283
What method is suggested for improving predictive accuracy in clinical models?
Shrinkage and penalized likelihood
## Footnote
These methods help in enhancing model performance.
284
What is the main contribution of the study by Putter H., Fiocco M., Geskus RB?
Tutorial in biostatistics on competing risks and multi-state models
## Footnote
This tutorial provides foundational knowledge for analyzing complex survival data.
285
What is the 'multiple imputation, then deletion' method used for?
Handling missing outcome data
## Footnote
This method has been evaluated for bias and precision in epidemiological studies.
286
What is the purpose of QRISK3 risk prediction algorithms?
Estimate future risk of cardiovascular disease
## Footnote
These algorithms were developed and validated through a prospective cohort study.
287
What does calibration refer to in predictive analytics?
The agreement between predicted probabilities and observed outcomes
## Footnote
Calibration is often considered a critical aspect of model evaluation.
288
True or False: Variable selection methods guarantee improved performance in clinical prediction models.
False
## Footnote
Regression shrinkage methods do not always lead to better predictive accuracy.
289
Fill in the blank: _____ is a method for evaluating clinical prediction models that involves comparing predicted and observed outcomes.
Calibration
## Footnote
Calibration techniques are essential for assessing model reliability.
290
What is 'data-driven subgroup identification' in clinical trials?
A method for analyzing specific subgroups within a larger population
## Footnote
This technique helps tailor interventions based on subgroup characteristics.
291
What does the study by Riley RD, Collins GS focus on?
Stability of clinical prediction models using statistical or machine learning methods
## Footnote
Understanding stability is vital for the implementation of these models.
292
What is the main finding from the study by Van Calster B., Steyerberg EW regarding prediction models?
There is no such thing as a validated prediction model
## Footnote
This highlights the ongoing challenges in model validation.
293
What does the term 'dynamic prediction systems' refer to?
Continual updating and monitoring of clinical prediction models
## Footnote
These systems adapt to new data over time for improved accuracy.
294
Who are the authors of the paper discussing the evaluation of clinical prediction models?
Collins GS, Dhiman P, Ma J, et al.
## Footnote
This paper emphasizes the need for thorough evaluation from development to validation.
295
What is the purpose of decision curve analysis in clinical settings?
To evaluate prediction models, molecular markers, and diagnostic tests
## Footnote
This approach assesses clinical utility and benefits of different predictive strategies.
296
What does the Heckman selection model address?
Dealing with missing data in epidemiological studies
## Footnote
This model provides a framework for understanding selection bias.
297
What is the focus of the study by Efron B., Tibshirani R on the bootstrap method?
Improvements on cross-validation
## Footnote
The 632+ Bootstrap Method enhances model validation techniques.
298
What is the focus of the paper by B and Steyerberg EW?
Net benefit approaches to the evaluation of prediction models, molecular markers, and diagnostic tests
299
Which organization published guidelines on cardiovascular disease risk assessment and reduction?
National Institute for Health and Care Excellence (NICE)
300
What is the title of the paper by Vickers AJ and Elkin EB that discusses a method for evaluating prediction models?
Decision curve analysis: a novel method for evaluating prediction models
301
Who authored a step-by-step guide to interpreting decision curve analysis?
Vickers AJ, van Calster B, Steyerberg EW
302
What does the extension to decision curve analysis aim to evaluate?
Diagnostic tests, prediction models, and molecular markers
303
What machine learning method is discussed in the paper by Breiman L?
Random Forests
304
Who proposed a unified approach to interpreting model predictions?
Lundberg S, Lee S-I
305
What fallacy is described by Westreich D and Greenland S?
The table 2 fallacy
306
Who introduced the concept of causal inference?
Pearl J
307
What is the purpose of the guide authored by Bonnett LJ and colleagues?
Presenting clinical prediction models for use in clinical settings
308
What R package is mentioned for creating web applications?
shiny
309
What topic does the JAMA article by McGinley MP and colleagues review?
Diagnosis and treatment of multiple sclerosis
310
What are the key aspects covered in the paper by Ghasemi N and colleagues regarding multiple sclerosis?
Pathogenesis, symptoms, diagnoses and cell-based therapy
311
What does Goldenberg MM's review focus on?
Multiple sclerosis
312
What approach is discussed by Crayton HJ and Rossman HS for managing multiple sclerosis symptoms?
A multimodal approach
313
What was developed and validated by Chalkou K and colleagues related to multiple sclerosis?
A prognostic model for relapse in relapsing-remitting multiple sclerosis
314
What is the focus of the Swiss Multiple Sclerosis Cohort-Study (SMSC)?
Investigation of key phases in disease evolution and new treatment options
315
What is CaliForest in relation to health data?
Calibrated Random Forest for Health Data
316
What topic does Platt J's work address?
Probabilistic outputs for support vector machines
317
What do reporting guidelines aim to improve according to Moher D?
Doing better for readers
318
Fill in the blank: The paper discusses the evaluation of _______ models.
prediction
319
What are deterministic models based on?
Precise inputs that produce the same output for a given set of inputs.
320
What do stochastic models incorporate into the modeling process?
Randomness and uncertainty.
321
List the pros of deterministic models
* Transparent cause-and-effect relationship
* Computationally efficient
* Require less data for accurate predictions
322
List the cons of deterministic models
* Assume all variables are known and accurately measured
* Do not account for uncertainty and randomness
323
List the pros of stochastic models
* Consider uncertainty and randomness
* Offer a range of possible outcomes
324
List the cons of stochastic models
* Demand more extensive data and computational resources
* More complex to interpret
325
What type of scenarios are deterministic models suitable for?
Well-defined and predictable inputs and outputs.
326
What type of scenarios are stochastic models suitable for?
Situations where the future is uncertain and unpredictable.
327
What do deterministic models assume about variables?
That all variables are known and can be accurately measured.
328
What do stochastic models allow decision-makers to assess?
The likelihood of different outcomes.
329
How do deterministic models relate to machine learning?
They aim to find a fixed relationship between inputs and outputs.
330
How do stochastic models relate to machine learning?
They incorporate randomness and uncertainty, capturing complex patterns.
331
In what areas do deterministic models excel?
Scenarios with clear cause-and-effect relationships.
332
In what areas do stochastic models excel?
Scenarios where the future is uncertain and unpredictable.
333
What is deterministic risk assessment based on?
Fixed inputs and assumptions.
334
What does stochastic risk assessment incorporate?
Randomness and uncertainty into the risk analysis process.
335
What do deterministic models require in terms of data?
Less data compared to stochastic models.
336
What do stochastic models require in terms of data?
A larger dataset to capture randomness and variability.
337
Fill in the blank: Deterministic models provide a clear _______ relationship between inputs and outputs.
cause-and-effect
338
Fill in the blank: Stochastic models provide a range of possible _______.
outcomes
339
True or False: Deterministic models are more complex to interpret than stochastic models.
False
340
True or False: Stochastic models can provide more accurate predictions when randomness is present.
True
341
What is a key takeaway regarding deterministic models?
They may oversimplify real-world complexities.
342
What is a key takeaway regarding stochastic models?
They account for uncertainty but require more resources.
