Training Index

This page gives a complete index of everything I currently deliver within John varlow | Training and Consultancy  courses.  The index will get bigger as we deliver more training courses.

It is the intention that each of the items in the index will have a description of what is covered within the training but this is a work in progress so please bear with us as this is developed.

A

What is absolute risk, how do you calculate it, and why do we use it?
What is absolute risk reduction? How does it relate to absolute risk? How is is calculated, and why do we use it?
What is an abstract? What should go into an abstract? When should an abstract be written?
What is an abstract for a presentation? How do you write one? What are the rules?
A key part of data quality.  How do you ensure data is accessible? What are the problems?
A key part of data quality. How do you ensure your data is accurate? What are the problems?
Error, trend and seasonality models following an additive approach for errors. What are they and why are they chosen?
A project management technique often used in software development. What is an agile project? What are the benefits and when should they be used?
Time series data is used a lot in the NHS and Social Care.  How do we decide which of our ARIMA models fits the data best? How does the AIC help with this? What is the difference between AIC and BIC?
The level of chance set within studies, to decide sample size and statistical certainty. What is alpha and how do we set it appropriately?
What is a hypothesis? What is special about the alternate hypothesis and how should it relate to your analytical / research question?  Alternative to what?
How do you perform analysis within SPSS?  What are the analysis menus? How do you interpret your results? (also delivered in PSPP if SPSS not available)
How do you determine how you are going to analyse your data before you have collected it? How should the analysis strategy be written
How is data collected and grouped? What is the difference between an independent measure and a repeated measure design?
Why is asking the right question important? How do you get people to ask the right question? What should go into a good question? How does the question relate to how you will do your analysis?
What is outcome measurement? What are the different approaches?  How do individual outcomes differ to research or population level outcomes?
Statistical power is important within research. It is related to sample size, play of chance and clinical differences.  What is power, and when should an a-priori power be calculated?
What is an ARIMA model? What is meant by an ARIMA(p,d,q)? What is meant by an ARIMA(p,d,q),n,(P,D,Q)? When shouldn't an ARIMA model be used?
How do you pick the right ARIMA model? How do you ensure your data is stationary? How do you use ACF and PACF?
How do you interpret the results / output of your ARIMA analysis. How do you determine whether you have a good model? How do you predict future values?
Sometimes variables seem associated. How can you test this and what does it mean?
What makes a quality indicator? What tools are available to have your indicator assured independently?
How do you define audit? Why do people define it incorrectly? How does audit relate to research, and what are the key differences?
Audits go through a lifecycle.  What are the key components of the audit lifecycle?
The ACF is used within ARIMA models to determine what model to use.  How do we do this?
ARIMA models are ways of looking at time series data. What is the autoregressive element in an ARIMA model. How does it differ from normal regression?
We use the word 'average' all the time and we often use it to refer to the arithmetic mean. What other averages do we use? Which is the best average given the type of data that we have?
There are a number of simple time forecasting methods. One of these is the average method?  What is the average method and how is it calculated?

B

What is a bar chart? How do they differ from histograms? What sort of data should we use in a bar chart?
What should be included in the background section of a  research proposal? How does this relate to the introduction? Where does the literature review fit?
When creating dummy variables for regression models you always need to specify a baseline category. What is a baseline category? How do you pick which category is the baseline, and how do you interpret your results?
A high level overview of the BATS method of time forecasting, currently only available within the statistical software R
Time series data is used a lot in the NHS and Social Care.  How do we decide which of our ARIMA models fits the data best? How does the AIC help with this? What is the difference between AIC and BIC?
What is the bell shaped curve? Why is it important? What other names do we have for it? How does it relate to other distributions?
High level overview of the distribution, its link to the binomial distribution and how it relates to other distributions
Statistical power is necessary in all research and analysis.  What is power and how does this relate to B.  How do we set B and how does it relate to sample size, statistical and clinical significance?  How does the concept of power relate to sensitivity, specificity, type I and Type II errors?
All human research has bias. What is bias? What are the different types and how can we minimise it?
High level overview of the distribution, its link to the Bernoulli distribution and how it relates to other distributions. What is its role in probability?
Regression methods are a type of multi-variate analysis. Why do we use regression methods? What is logistic regression and how does it relate to linear regression? How do we interpret results as both odds ratios and probabilities?
What is bi-variate analysis? How does it differ from uni-variate or multi-variate analysis? What sort of bi-variate techniques are there?
Blinding is essential in RCTs, and controls for the placebo effect.  What is blinding, and how do we achieve it? What is the difference between single and double blinding? What are some of the pitfalls?
Otherwise known as a Forrest plot.  What is it and how is it used in systematic reviews and meta-analysis? What are the implications for non-significant results?
There are many different types of randomisation. What is block randomisation? How do you define blocks? When is it used?
Body language is key when presenting to / training your audience. What are the dos and donts of body language?  How can you change what you do?
One of a number of post-hoc tests. Bonferroni is used following analysis of variance (ANOVA).  Why are these types of tests needed?  How do they relate to t-tests?
What is a box and whisker plot? What sort of data should be put in this kind of plot? What are the strengths of this plot? Why aren't they used more than they are?
Data often has to be transformed, particularly when doing regression or time forecasting. What is the Box Cox transform and how does it relate to power and log transforms?
Autocorrelation often occurs n data series, particularly time series.  What is the Breusch Godfrey test and how does it relate to autocorrelation?

C

All processes have variation, both natural and special.  Given this, how capable is your process in meeting a particular target and how do you measure this?
What is case-control design?  When is it used and what are the strengths and weaknesses of such an approach?
What is case series?  When is it used and what are the strengths and weaknesses of such an approach?
Case study design is one of the 3 types of research design. What is it~? How is it carried out, and what are the strengths and weaknesses of this design approach?
Any type of data can be put into categories. What are independent categories and related categories? What is the difference between ordered and un-ordered categories?
There are two types of categorical data, these are nominal and ordinal data. What is special about categorical data? How should it be treated and how does it relate to the tests that can be performed on the data?
Many pieces of research and analysis try to prove that there is causation i.e. one factor causes another.  What are the problems with this approach? What are the research designs that support causation? What are the implications of getting it wrong?
This theorem is one of the fundamentals of inferential statistics and underpins the concepts of standard error and confidence intervals. How does it relate to the distribution of your data? How does it build on the concepts of normal distributions and standard deviation?
What is the centroid? How does it relate to the line of best fir within linear regression?  How is it related to summary statistics and mean values?  
We often treat our results (of research or analysis) as facts.  But are we really certain about anything?  What is the role of certainty in statistics? How certain can we really be?
No matter how good our research or analysis, sometimes strange things happen for no reason.  How do we account for chance effects within our analysis?  What is the link between chance effects and significance?
Any analysis has to take into account the play of chance. Chance is instrumental in determining sample size. How does setting alpha reflect on confidence intervals and statistical certainty?
There are a number of ways to present our data. These mostly use a selection of charts.  How do you know what chart to use? How does the choice of chart relate to the type of data that you have?
Based on the chi-square distribution. What is the chi-square test? How many independent groups can it be used on? What is the role of Fisher's test? How do you calculate by hand?
Much of the data that we deal with (particularly in the NHS and Social Care) is classification data.  What is classification data? How does it differ from terminology? What examples do we have? How do we analyse this type of data?
We often talk bout significance, but what do we mean? What is clinical significance? How does it differ from statistical significance? How does it impact on sample size calculations and power?
There are a number of different terminologies currently within systems. CTV3 is one such terminology. How is it used and what are the future plans?
There are a number of ways of identifying your sample.  What is cluster sampling? What are the strengths and weaknesses? How does it compare to other sampling methods?
Cochrane's Q is a non-parametric test for related groups.  How is it calculated? What distribution is it based on? How many groups can it be used on? How does it relate to the McNemar test?
What is a cohort analytic study?  When is it used and what are the strengths and weaknesses of such an approach?
What is cohort design?  When is it used and what are the strengths and weaknesses of such an approach?
We often have a lot of data to present, and combination charts, such as a clustered bar are commonly used. What is the problem with these types of charts? When should they be used? Are there alternatives?
The word comparison is used to explain groups of subjects within research studies and forms part of the PICO approach. What is the relationship between a control group and a comparison group?
A key part of data quality.  How do you ensure data is complete? What are the problems?
How do you perform complex data manipulation within SPSS?  What are the manipulation menus? What can go wrong? (also delivered in PSPP if SPSS not available)
Measures are usually simple or compound. What is a compound measure and why are they used? What are the problems with compound indicators?
P-values have been used to express statistical probability for many years. However, it is now good practice to use confidence intervals. What are confidence intervals? How do you calculate them? What advantage do they have over p-values?
There are some variables that cause problems when looking at associations and causal relationships. What are confounding variables? How do they relate to other variables? What can you do to remove their effect?
Research should always consider and mitigate for ethical issues. How does the conservation of resources relate to ethics and how can you mitigate for this? 
Many statistical models contain a constant. What is a constant and how do you interpre them within regression and forecasting models?
All data is either categorical or continuous. What is different between categorical and continuous data? How should we treat categorical data?  What do we mean by interval and ratio data? What is the relationship between continuous and ordinal data, and when should you treat continuous data as if it is ordinal?
Control is one of the most important aspects of project management. What is control?  Why is t so important and how do you ensure it takes place?
The word control is used to explain groups of subjects within research studies. What is control and why are some groups inappropriately named.  What is the relationship between a control group and a comparison group?
Control limits are used with control charts. What are they, and how do they differ from confidence intervals?
There are a number of ways of identifying your sample.  What is convenience sampling? What are the strengths and weaknesses? How does it compare to other sampling methods?
How do you see how variables are associated? What are correlations and how do you interpret them? Which correlation techniques are associated with different types of data?
What is correlation? What tests should you use? How do you interpret specific coefficients. How do you calculate them?
What is a hazard and hazard ratio? How does this differ from risks and risk ratios?  What role does time play in hazards?  How are Kaplan Meier survival curves related to proportional hazards?
What is the difference between appraising and reading an article? What do we mean by information overload?  How do we keep on top of the literature? Where did critical appraisal originate? What tools are there?
What is the process to follow when critically appraising a paper? What is the most important part of a paper. How do you decide on the relevant tool? What screening questions should be asked before spending time on a paper?
What is CASP? Where was it started? What tools have they made available for critical appraisal?
Many studies are designed as being cross sectional. What does this mean? How are cross sectional studies established and what conclusions can be drawn from them? What are their strengths and weaknesses?
All time series data have a number of different properties. One of these is cyclicity. What is cyclicity and how does it vary from seasonality? 

D

Forecast data that follows a linear trend often over estimates future forecast values. One way to account for this is damping, or dampening. What is damping? How is it calculated and applied to ETS models?
All changes of data standards within the NHS in England are included within the NHS Data Model and Dictionary. What is the dictionary? How is it updated, and how are system suppliers and the NHS service notified of changes? What is the implication of not following the dictionary?
Research should always consider and mitigate for ethical issues. How does data protection relate to ethics and how can you mitigate for this? How has data protection changed with the DPA 2018 and GDPR?
Data quality is essential if our analysis is going to be meaningful. What are the elements of data quality? How do you ensure your data is as high quality as possible?
A key part of data quality.  How do you ensure data is accessible? What are the problems?
A key part of data quality.  How do you ensure data is accurate? What are the problems?
A key part of data quality.  How do you ensure data is complete? What are the problems?
A key part of data quality.  How do you ensure data is relevant? What are the problems?
A key part of data quality.  How do you ensure data is accessible? What are the problems?
How is data coded within systems? How do we ensure that data is collected in the same way and that systems can talk and connect to one another? What is interoperabilty? What is the data dictionary and how do data standards get in there?
Data often has to be transformed, particularly when doing regression or time forecasting. What are the different types of transforms?
How do you define audit? Why do people define it incorrectly? How does audit relate to research, and what are the key differences?
There are two categories of statistics, descriptive and inferential. What are descriptive statistics@ When do you use them and what do they tell you? What are the limitations of descriptive statistics?
How do you define evaluation? How does evaluation relate to research and audit?
What do we mean by an indicator.  How is it different to a metric?  What do we mean by the underlying construct of the indicator?
There are two categories of statistics, descriptive and inferential. What are inferential  statistics? When do you use them and what do they tell you? What are the limitations of inferential statistics?
Metrics and indicators are used extensively within the NHS and Social Care.  What is the difference between a metric and an indicator?  How do we use metrics and how are they misused?
Numbers are everywhere.  We use numbers to count things. But what are numbers? Can we always treat them in the same way?
How do you define research? Why do people define it incorrectly? How does audit relate to research, and what are the key differences?
What is service development? How is it different to evaluation and why does it need distinguishing from research or audit?
Time series data is used across the NHS and Social Care. What is it, and how can we analyse and forecast from it?
Degrees of freedom are fundamental to all statistical tests. What do we mean by independent observations and how does this relate to degrees of freedom? What statistics do we commonly use that incorporate degrees of freedom?
Often used in judgemental forecasting. What is the delphi method and what are its limitations?
Numerators and denominators are the primary ingredients of rates. What are they, and what is the difference between the two?  Do numbers in your numerator always have to be in your denominator? 
What are dependent variables? How do they differ from independent variables?  How do we use dependent variables?  Are their any other variables we need to consider?
There are two categories of statistics, descriptive and inferential. What are descriptive statistics@ When do you use them and what do they tell you? What are the limitations of descriptive statistics?
What is case-control design?  When is it used and what are the strengths and weaknesses of such an approach?
What is case series?  When is it used and what are the strengths and weaknesses of such an approach?
Case study design is one of the 3 types of research design. What is it~? How is it carried out, and what are the strengths and weaknesses of this design approach?
What is cohort design?  When is it used and what are the strengths and weaknesses of such an approach?
Many studies are designed as being cross sectional. What does this mean? How are cross sectional studies established and what conclusions can be drawn from them? What are their strengths and weaknesses?
One of the three main study designs. What is experimental design? Why is it seen to be the best design? What are its limitations?
One of the three main study designs. What is survey design?  What are the strengths and weaknesses of such a design? 
Part of the scientific method. What is determinism? Why is it important in research and analysis? What  other ways of thinking are there?
Deviation is a concept used in a number of statistics, the most common being standard deviation. What is deviation and how do we measure it? How does deviation link to degrees of freedom?
What is dichotomous data? What type of data is dichotomous data?  What can we do with dichotomous dependent variables? Can we treat dichotomous data as continuous data?
Time series data needs to be stationary before an ARIMA model can be built.  How do we make our data stationary?  What is differencing and how does it make our data stationary? 
Dissemination is a key part of the research process.  What methods of dissemination are there? Are some methods better than others? Does the audience play a part in deciding which method you use?
Calculated rates and ratios are often subject to different population structures. How can we adjust our data so that rates are comparable between populations. Can we use national population structures to adjust our data. Can this always be done? What should you standardise for? What is the difference between direct and indirect standardisation?
There are many types of data. What is discrete data? How does it differ from integer continuous data? How should it be treated, and what statistical tests can be used?
What do we mean by dissemination and why should we disseminate our research findings?
What is a dissemination strategy and why should you have one when you are doing research?
All data is distributed in particular ways. What are the particular types of distribution and how do they relate to each other? Why do we concentrate on the normal distribution?
High level overview of the distribution, its link to the binomial distribution and how it relates to other distributions
High level overview of the distribution, its link to the Bernoulli distribution and how it relates to other distributions. What is its role in probability?
High level overview of the Chi Square distribution and how it relates to other distributions
High level overview of the exponential-distribution and how it relates to other distributions
High level overview of the F-distribution and how it relates to other distributions
High level overview of the Gaussian distribution. Is it different to the normal distribution and how does it relate to other distributions
What is a leptokurtic distribution and how does it relate to kurtosis?
What is a mesokurtic distribution and how does it relate to the normal distribution
High level overview of the normal distribution, its other names, how it differs from the Standard Z and how it relates to other distributions
What is a platykurtic distribution and how does it relate to skewness?
High level overview of the poisson distribution and how it relates to other distributions
High level overview of the student's t-distribution, its link to the normal distribution and how it relates to other distributions
High level overview of the distribution, its link to the normal distribution and how it relates to other distributions
High level overview of the distribution and how it relates to other distributions
Ratio data is a special type of continuous data. What is the doubling up principle and how does it relate to ratio data?
There are a number of simple time forecasting methods. One of these is the drift method?  What is the drift method and how is it calculated?
We create dummy variables for regression models when using categorical data. How do you create dummy variables? How many dummy variables are there? What is a baseline category? How do you pick which category is the baseline, and how do you interpret your results?
Autocorrelation often occurs n data series, particularly time series.  What is the Durbin Watson test and how does it relate to autocorrelation?

E

There are many types of validity when considering results. What is ecologic validity? Why is it important and why is it often overlooked?
Time Series data is used a lot within the NHS and Social Care. What are the elements of time series data and how do they differ from each other?
Part of the scientific method. What is empiricism? Why is it important in research and analysis? What  other ways of thinking are there?
What is epidemiology? When is it used, and what are the major methods within it?  How does epidemiology relate to research and analysis?
What does it mean for samples to have equal variance?  What do you do if this is not the case? How does this relate to the Z-test and the T-test for two independent groups. What is Levene's statistic?
Much of inferential statistics is concerned with error.  What is sampling error and how can it be used to our advantage?
Much of inferential statistics is concerned with error.  What is sampling error and how can it be used to our advantage?
We talk a lot about ethics within research.  Research studies need ethical approval before studies can start. What is ethical approval? What is considered and how long does it take?
What are ethics in research? What are the ethical issues that you need to consider? How do you mitigate for these?
Research should always consider and mitigate for ethical issues. How does the conservation of resources relate to ethics and how can you mitigate for this? 
Research should always consider and mitigate for ethical issues. How does the data protection relate to ethics and how can you mitigate for this? 
Research should always consider and mitigate for ethical issues. How does informed consent relate to ethics and how can you mitigate for this? 
Research should always consider and mitigate for ethical issues. How does minimising discomfort relate to ethics and how can you mitigate for this? 
Research should always consider and mitigate for ethical issues. How does the protection of participants relate to ethics and how can you mitigate for this? 
Research should always consider and mitigate for ethical issues. How do the taboos of the community relate to ethics and how can you mitigate for this? 
What is ethnography? How does it relate to the interpretevist paradigm? 
What are ETS models? What does ETS stand for, and when are these models used?  How do ETS models relate to other forms of time series forecasting models?
How do you define evaluation? How does evaluation relate to research and audit?
What are event rates? How are they calculated? How are they used and what are the problems with using them?
One of the three main study designs. What is experimental design? Why is it seen to be the best design? What are its limitations?
What is exploratory analysis? When should you do it?  What does it tell you?
High level overview of the exponential-distribution and how it relates to other distributions
Why do we use smoothing methods? What is exponential smoothing? What methods are included in exponential smoothing? How does it relate to ETS models?
Often (if incorrectly) refereed to as ETS, what is exponential triple smoothing?  What are the properties of time series data that are addressed using triple smoothing?
All data follows trends. What is an exponential trend? How does it relate to growth? Is it the same as a polynomial trend?
There are a number of different types of reliability. What is external reliability? How does it differ from internal reliability? How does inter and intra observer reliability fit in with external reliability?

F

High level overview of the F-distribution and how it relates to other distributions
Feasibility is key to a successful research project. What is feasibility and how can you ensure that your project is feasible?
Finance / resource considerations are key in delivering a successful research project. How do you ensure that you know what resources and finance you will need? What is the connection between finance / resource and timescale of a project?
Most nominal data is compared using tests based on the chi-square test.  When should Fisher's test be used instead? What are the rules?
Microsoft Excel has a number of inbuilt statistical functions that keep being developed with each release.  What is the FORECAST function? How did it start off in MS Excel and what can you do with the function in the latest version.  How does the FORECAST function differ from the TREND function, and what is the link with linear regression and exponential smoothing? 
A lot of data within the NHS and Social Care is time series data.  How do you forecast what might happen in the future given historical data? What forecasting models can you use?
Otherwise known as a blobbogram (by CASP).  What is it and how is it used in systematic reviews and meta-analysis? What are the implications for non-significant results?
Named after the French mathematician, what is a Fourier series? How is it used within regression models?  How does it relate to seasonality?
What is the Friedman test? When do you use it? How does it relate to the McNemar test?
Made famous by Walter Shewhart, what are funnel plots? When are they used and what do they tell you?

G

We construct indicators and metrics on a regular basis, introduce new initiatives and set targets.  What s the problem with doing this? What is gaming and perverse incentives? How can we avoid the issue?
What tools do we have in project management? How do we record what the project plan is and how we are meeting key milestones? What is a Gantt chart and why are they useful?
High level overview of the Gaussian distribution. Is it different to the normal distribution and how does it relate to other distributions
A linear regression line is made up of a constant and a gradient.  How do we determine what the right gradient is? How do we calculate it?
There are a number of ways to graphically present our data. These mostly use a selection of charts.  How do you know what chart to use? How does the choice of chart relate to the type of data that you have?
Grounded theory is a key concept within qualitative research. What is grounded theory and how is it used?
The GROWTH function within MS Excel is a key statistical function looking at forecasting. How does the GROWTH function link to polynomial regression? How is it used appropriately?

H

Hand gestures are used a lot in presentations, for effect and emphasis. How can you use hand gestures to best effect?  What are some gestures that you should avoid using?
The Hawthorne Effect occurs in prospective studies. What is the Hawthorne effect and can you control for it? 
Healthcare Resource Groups (HRGs) are used within the NHS for costing purposes. What are HRGs? How are they defined and used? 
Heteroskedasticity is a key part of regression methods.  What is heteroskedasticity and how does it relate to residuals and autocorrelation? 
What is a histogram? How do they differ from bar charts? What sort of data should we use in a histogram?
What is exponential smoothing and how does Holt's linear method relate to Brown's Method and Holt Winter's method?  When should each method be used and how do they relate to seasonality? Is there a different method when relationships are not linear?
What is exponential smoothing and how does Holt Winter's method relate to Brown's Method and Holt's linear method?  When should each method be used and how do they relate to seasonality? Is there a different method when relationships are not linear?
What do we mean by homogeneity of variance? Why is it important, and how is Levene's statistic used to interpet results where various isn't homogenous?
Homoskedasticity essential in regression methods. What is it, and how does it relate to heteroskedasticity?  How can you make your residuals homoskedastic?
When  should you write a hypothesis?  How does this relate to your research / analytical question?  What is an alternate hypothesis and what is it the alternative to?
When  should you write a hypothesis?  How does this relate to your research / analytical question?  What is a one-tailed hypothesis and how does this relate to one tailed statistical tests?
When  should you write a hypothesis?  How does this relate to your research / analytical question?  What is a null hypothesis? How does it relate to the alternate hypothesis? Can you prove a null hypothesis? Why formulate a hypothesis that hypothesises nothing?
What do we mean by a statistical hypothesis?  Is this different to a research hypothesis? What do we mean when we talk about significant differences in a statistical hypothesis?
What is the difference between a one-tailed and a two-tailed hypothesis?  How does this affect the analysis that is performed.  Are there any rules as to which type of hypothesis is picked?

I

What are ICD codes? How are they used in practice and why should we care? What is the difference between different versions? How are they analysed, and how do they relate to other coding structures? What do we mean by classifications and how does this differ from terminology?  What do we mean by coding granularity?
What do we mean when we say that a solution has been implemented?  Is this the end of the process or is there more to do? What is a post implementation review and what do we do with it?
What is the difference between incidents and incidence?  How do we define incidence, and why do most people calculate it incorrectly?  What do we mean by populations at risk?  How does incidence relate to prevalence and why is it important to understand your outcome event? What should be quoted alongside incidence rates? 
What is an incident case?  Does everyone have the same risk? How do incident cases relate to incidence rates?
What are independent samples?  How do they relate to research / analytical designs?  What sort of statistical tests can be used on independent samples?  What needs to be taken into account?
What is the independent samples t-test? When should it be used? What do we men by independent samples?  Is this test always appropriate? What is the t-distribution and how does it relate to the normal distribution?  Why is it sometimes called the Student t-test?
What is an independent variable? What sort of models are we looking at if we are talking about independent and dependent variables? How is a dependent variable different?  How do these relate to confounders?
What do we mean by an indicator.  How is it different to a metric?  What do we mean by the underlying construct of the indicator?
How do you construct an indicator? What are the factors that need to be taken into account?  Are complex indicators better than simple indicators?
Calculated rates and ratios are often subject to different population structures. How can we adjust our data so that rates are comparable between populations. Can we use national population structures to adjust our data. Can this always be done? What should you standardise for? What is the difference between direct and indirect standardisation?
What do we mean by inference?  What are the problems associated with it? How can we use statistics to give some certainty around inference?
There are two categories of statistics, descriptive and inferential. What are inferential  statistics? When do you use them and what do they tell you? What are the limitations of inferential statistics?
What are influencing variables in research or statistical models?  Are they independent variables or dependent variables? How do they relate to confounders?
What is an information standard? Why are they necessary? How do they relate to interoperability?  Are they different to data standards?
What do we mean by informed consent. How is this different to explicit consent? When do you need informed consent? How does this relate to data protection rules?
Time series data needs to be stationary before an ARIMA model can be built.  How do we make our data stationary?  What is differencing and how does it make our data stationary? How does an integrated model relate to differencing?
What are intellectual property rights. How does IPR differ from copyright and patents? What needs to be done to ensure IPR isn't violated? Does IPR have monetary value?
Intention to treat is crucial in research studies, particularly randomised controlled trials.  What do we mean by intention to treat and why is it violated so often. What is the implication of intention to treat being violated? What can we do about missing observations? What do we mean by last observation carried forward?
How can we measure dispersion, or spread of data?  Do we have similar measures for different types of data?  If not, what  is the inter-quartile range and what data should it be used with?
Our data is stored in numerous systems, whether health, education financial etc. Often data needs to be transferred from one system to another.  What are the problems with this? Are there any risks and how do we mitigate for these? 
There are many different regression models, all of which have different output when using statistical software. How do we interpret this, and what are the similarities and differences of different models?
There are two different  paradigms within research.  These are the positivist and the interpretevist paradigm.  What are these, and what is the difference between them.  How do we decide which paradigm to follow?
All data is either categorical or continuous. What is different between categorical and continuous data? How should we treat categorical data?  What do we mean by interval and ratio data? What is the relationship between interval and ordinal data?
Many research studies are based around an intervention. What is an intervention and what do we need to consider before the effect of an intervention is assessed?
What is an intervention group within a Randomised Controlled Trial. How should an Intervention group be treated, and what is its relationship to the comparison group? How does randomisation, blinding and intention to treat ensure that an intervention group is analysed appropriately
What is reliability and why is it important? What is the difference between inter-observer and intra-observer reliability? How does it relate to the concepts of internal and external reliability?
What is reliability and why is it important? What is the difference between internal and external reliability? How does it relate to the concepts of inter-observer and intra-observer reliability?
What is reliability and why is it important? What is the difference between inter-observer and intra-observer reliability? How does it relate to the concepts of internal and external reliability?
What is the purpose of an introduction within a research proposal. What sort of things should be included in the introduction and how does it differ from the background?

J

A lot of data within the NHS and Social Care is time series data.  How do you forecast what might happen in the future given historical data? What forecasting models can you use? What is judgemental forecasting, and how does it differ from other forecasting methods?  How does it relate to the Delphi method?

K

What are Kaplan Meier survival curves?  When should they be used? How do they relate to Cox's proportional hazards? Is the Kaplan Meier curve able to be calculated manually?
What is correlation? What tests should you use? How do you interpret specific coefficients. How do you calculate them? When should you use Kendall's Tau Correlation? How is it related to Spearman's correlation coefficient?
What is knowledge, and why is it important to determine the type of knowledge we have. How does this relate to the scientific method? Why shouldn't we use private knowledge to inform our practice?
What is knowledge, and why is it important to determine the type of knowledge we have. How does this relate to the scientific method?  What is the difference between professional knowledge and scientific knowledge?
What is knowledge, and why is it important to determine the type of knowledge we have. How does this relate to the scientific method? What is the difference between public knowledge and professional knowledge? 
What is knowledge, and why is it important to determine the type of knowledge we have. How does this relate to the scientific method? What is scientific knowledge and why is it the only knowledge that should inform practice?
What is the kruskal wallis test? When should it be used?  What is the comparative test for two groups? What are the assumptions within the test?

L

What do we mean by lag when we are talking about time series data? Why is it necessary to include lag in forecast models and which models best minimise the play of lag?
What do we mean by last observation carried forward?  How does this relate to the idea of intention to treat? 
Least squares is a method used in a number of statistical tests including regression.  What is the method of least squares and why is it important?

M

N

O

P

There are two different  paradigms within research.  These are the positivist and the interpretevist paradigm.  What are these, and what is the difference between them.  How do we decide which paradigm to follow?
Statistical power is important within research. It is related to sample size, play of chance and clinical differences.  What is power, and when should a post-hoc power be calculated?
Statistical power is important within research. It is related to sample size, play of chance and clinical differences.  What is power, and when should an a-priori power be calculated.  How does this relate to  post-hoc power?

Q

R

S

Statistical power is important within research. It is related to sample size, play of chance and clinical differences.  What is power, and when should an a-priori power be calculated? How does this relate to a post-hoc power?

T

A high level overview of the TBATS method of time forecasting, currently only available within the statistical software R
Much of the data that we deal with (particularly in the NHS and Social Care) is classification data.  Yet we ar moving toward recording information using terminology (namely Snomed CT).  What is terminology data? How does it differ from classifications? What examples do we have? How do we analyse this type of data?
One of a number of post-hoc tests. Bonferroni is used following analysis of variance (ANOVA).  Why are these types of tests needed?  How do they relate to t-tests?
Autocorrelation often occurs n data series, particularly time series.  What is the Breusch Godfrey test and how does it relate to autocorrelation?
Based on the chi-square distribution. What is the chi-square test? How many independent groups can it be used on? What is the role of Fisher's test? How do you calculate by hand?
Cochrane's Q is a non-parametric test for related groups.  How is it calculated?  What distribution is it based on? How many groups can it be used on? How does it relate to the McNemar test?

U

V

W

XYZ

0-9