Sep 14, 2024
Class 1
Reading: Nurse Killam Levels of Measurement: How to Remember Differences. https://www.youtube.com/watch?v=LPHYPX BK_ks Type of research affects data Highest level of data collection can help with best outcome Order of level of measurement - NOIR
Age - can only be measured by Nominal Level of income - measured on multiple level depending on how question is asked
Nominal - lowest level, discrete categories, no specific order, sounds like name - simply naming category, it is simply naming, not numerical Can fall into 2 levels - dichotomous ( things with 2 options like male & female, or yes or no) or categorical (more than 2 possible options) Allows 2 options of data collection Can’t find median/mean - since there is no specific order Presented as bar/pie chats Ordinal - has ordered categories Used to establish order Categories has natural order - numbers are arbitrary and assigned to categories - to establish ranking Very satisfied, dissatisfied scale i.e - on a scale of ranking from 1-10 Interval Similar like Ordinal - has categories, equal distance between values Ordered categories i.e the temperature scale Can do more data - median, mean Has arbitrary zero - thus can’t divide by zero Example, temperature scale, 79 and 80 difference is similar to 39 and 40 difference Ratio = zero Most precise, ordered, exact values, equal intervals, natural zero All operations are possible Only possible with physical measurements - height, weight,
To determine the level of measurement - decision tree Is it ordered? If no, it is nominal If yea, it is equally spaced? If no, ordinal If yes, does zero mean none? If no, it is interval If yes, it is ratio
Nominal has mode only because it has no specific numbers Ordinal has mode, median, can’t do mean since zero is non-existent Bar Charts, Pie Charts, Histograms, Stem Plots, and Time Plots https://www.youtube.com/watch?v=uHRqk GXX55I We use this to display data categorical data - bar charts, pie charts Quantitative data - stemplots, histograms, timeplots - to show information Bar charts - display the frequency on Y axis, and the value of the categorical variable on x axis
Pie charts - relative size of each value in relation to another Histogram - shows the distribution of the collected data
Relative frequency - represents proportion values of each interval in relative to the whole, do calculation Find the total sum of the frequency - here it is equal to 50 Divide each value by the sum - 16/50 = 0.32, can be converted to percantages
Stem plots - do I need to know this? Timeplot - distribution of time Joel Selanikio: The big data revolution in healthcare: https://www.youtube.com/watch?v=GHJRcMO w45k
Class Notes: Overview of statistics in healthcare Fundamental concepts. ● Types of data (quantitative vs qualitative) Quantitative - numerical data, two types ● Collecting data Categorical/discrete data Can be nominal, ordinal, categorized interval or ratio Continuous Can it be interval or ratio - why? ● Levels of Measurement Nominal Describes categories, labels, names For example, eye colour Ordinal Categories The order matters Rank or order exists Interval Evenly spaced Categories and order matter Examples ; temperature scale You can take an average Ratio Evenly spaced Has categories and order matters Can have a meaningful absence or zero You can take an average Examples related to nursing Nominal - blood types- AB,O Ordinal - pain scales (0=no pain…10=worst) Interval - core temperature measured in degree celsius Ratio - patient’s weigh in KG ● Presenting data. Textual Describes using texts 97% of text messages were opened Tabular Reporting using data table Graphically histogram No gaps in between the frequency, leads to bell shaped Line graphs Continuous data Change over time Scatter plot Continuous data Data points Plot of 2 variables
Textbook review: Chapter 1. Introduction to Statistics and Levels of Measurement Chapter 2. Presenting Data
Class 2
Reading: Understanding descriptive and inferential statistics https://www.youtube.com/watch?v=g1NkoiJWpA What is statistics? The science of collecting, analyzing and interpreting and presenting data. 2 categories; descriptive statistics, and inferential statistics Descriptive statistics; organizing, summarizing data using numbers and graphs Data summary - bar graphs, histograms, pie charts, et. shape of graph & skewness Measures of central tendency: mean,median, mode Measures of variability - range, variance and standard deviations inferential statistics; Using sample data to make an interference or draw a conclusion of the population Uses probability to determine how confident we can be that the conclusion we make are correct - confidence intervals, margins of error
Mean Median and Mode: Understanding and Calculating Measures of Central Tendency – Nurse Killam https://www.youtube.com/watch?v= gYTwio S4mbo Measures of central tendency include - mean, median, mode - fall on the same midline point on a distribution curve Normal distribution - when data are gathered from interval and ratio level measures and plotted on a graph it will resemble a normal curve Best measure depends on - whether data has normal distribution or not If data is not appropriately distributed certain measures may be better than others Mean = average
Most common, best known to describe the centre of the distribution, only for interval and ratio data Median - is the middle, can’t be used for nominal data
Mode = Most frequent
Measure of central tendency, can’t be used if all scores are different, there can be multiple modes, unstable, only one that works for nominal data, cannot be used for further calculations
Normal Distributions, Standard Deviations, – Nurse Killam https://www.youtube.com/watch?v=HnMG KsupF8Q
Normal distribution - concept that talks about how data will look once plotted
Allows probabilities to be calculated
Inferential statistics requires that data to be distributed normally
Key features of normal distribution;
Centered
Fixed score distribution
Unimodal - one peak in distribution
Symmetrical - bell shape curve
Standard deviation - a way to measure how much variation exists in a distribution
Low standard deviation - values are close to the mean
High standard deviation - values spread out over a large range
In a normal distribution, 34% of scores fall between the mean and 1 standard deviation above the mean, therefore, based on it’s symmetry, approximately 68% of scores fall between 1 standard deviation above and 1 standard deviation below the mean, approximately about 95% of deviation fall between 2 standard deviations above and 2 standard deviation below the mean, and approximately about 99.8% fall 3 % above and below the mean
Z scores are used to measure how many standard deviations above or below the mean a particular score is.
Not all data is normal; Modality Symmetry - independent from modality, something can be superimposed and be mirror image of the other thus symmetrical Peakedness - peakedness determines the modality 1 peak - unimodal Two peaks - bimodal 2 or more peaks - multimodal
Kurtosis = it measures of the bell of the curve is normal, flat or peaked Using fisher’s method of kurtosis, a normal distribution bell curve has 0 K=0 - mesokurtic k>0 - leptokutic - peaks sharply k<0 - platykutic - flattened, highly dispersed Class Notes:
Descriptive statistics Presentation, organization and summarization of data Graphical representation & tables Ca-founder - something that affects your result but you are ot necessarily investigating in your research ● Measures of central tendency - NOIR Nominal = central tendency- Mode Ordinal = central tendency- Mode, Median Interval = central tendency- Mean, mode, median Ratio = central tendency- mean, mode, median Graphs; Nominal - bar, pie Ordinal - bar, pie Interval - bar, pie, box plot, histogram Ratio - histogram, box plot Plotting of data; Histograms; Display of the frequencies Visual representation of the data Useful for large data sets Shape depends on the data mean=mode=median Positively skewed Negatively skewed Multimodal Bimodal ● Measures of variability Common measures of variability; Range - how far are scores are from the mean Highest value - smallest value Variance - average of the squared ( cause of symmetry and it has to be positive to make sure they don’t cancel each other out) - it is a measure of how spread out values are around the mean. It changes the unit to squared Small variance - most scores close to the average Large variance - scores are widely spread Standard deviation - square root of the variance, gives information on the amount of variation Interquartile range - divides the distribution into quarters Median falls at the 50th percentile Interquartile range = Q3-Q1
Box plot - whisker box
It cuts stuff to boxes
● Frequency distribution and histograms
Report MEAN = for central tendency & standard deviation - when distribution is reasonably symmetrical
Report MEDIAN = for central tendency & inter quartile range - with variability and non-symmetrical distributions
Descriptive statistics
Inferring - we are saying this might happen, this is what we think - making predictions, uses data to estimate, test hypothesis, and make predictions
Can infer range of values = confidence intervals
Can infer a point of estimate = single value
Key concepts;
Probability
Normal distribution
Population and sample
Standard scores (Z-scores)
Types;
Parametric
Random sample
Unbiased selection
Normal distribution
Homogeneity of variance
Non-parametric statistics
Use it when any of the other reasons don’t work
Error
Will always occur
Maybe because of data collection, data interpretation, sample
Textbook review:
Chapter 3. Descriptive Statistics, Probability and Measures of Central Tendency
Class 3 Reading: Central limit theorem https://www.youtube.com/watch ?v=Pujol 1yC1_A
Class Notes: Probability ● Concept of probability ● Probability distributions ● Z-scores and standardisation Sampling ● Sampling ● Population ● Central Limit theorem
Textbook review: Heavey: Chapter 5. Sampling methods Chapter 7. Sample size and Effect Size
Class 4
Reading: Bias https://www.youtube.com/watch ?v=5K1H g-pSY1A Confidence intervals https://www.youtube.com/watch ?v=tFWs uO9f74o
Class Notes: Bias and error ● Sources of error ● Confidence intervals and margin of error
Textbook review: Chapter 3. Descriptive Statistics, Probability and Measures of Central Tendency
Class 5 Reading:
Class Notes: Collecting Data Hypotheses testing ● Null and alternate ● Type 1 v Type 2 error
Textbook review: Chapter 7. Type 1 and 2 Error, Power
Class 6 Reading: Introduction to critical values (reference tables at back of textbook * we refer to these throughout the course* Visualizing type 1 and type 2 errors https://www.youtube.com/watch ?v=k80p ME7mWRM Validity and Reliability of measurement by Alyson Froehlich https://www.youtube.com/watch?v=VTH WQOuEfiM Class Notes: Level of significance ● One -tail and two-tail tests Clinical vs statistical significance Reliability and validity
Textbook review: Chapter 6. Generating the Research Idea Chapter 4. Intro and Sections on validity and Reliability
Class 7 Reading: Dependent and Independent samples https://www.youtube.com/watch?v=mWE 8PxPoYJY Parametric and non-parametric tests https://www.youtube.com/watch?reload= 9&v=pWEWHKnwg_0 Class Notes: Dependent and independent samples Parametric and non-parametric tests. Deciding on which statistical test
Textbook review:
Class 8 Reading: Class Notes: Comparing two groups ● Student’s t-test ● Mann Whitney ● Paired t-test ● Wilcoxon Rank ● Signs test ● Chi Square test for independence
Textbook review: Chapter 9. Student t-test Chapter 8. Chi-square
Class 9 Reading: Class Notes: ANOVA ● Assumptions ● Interpretation ● Post-hoc tests Correlation ● Assumptions ● Interpretation Spurious results
Textbook review: Chapter 10. Analysis of Variance (ANOVA) Chapter 11. Correlation Coefficients
Class 10 Reading: Class Notes: Regression ● Simple linear regression ● Coefficient of determination ● Logistic regression ● Multivariate analysis
Textbook review: Chapter 12. Regression Analysis
Class 11 Reading: Randomized control trial: https://www.youtube.com/watch?v=553Tx UhtEbk Cohort and case control designs: https://www.youtube.com/watch?v=sGfIK mKMRdg Numbers needed to treat https://www.medpagetoday.com/blogs/thi rdopinion/50273 Class Notes: Critical appraisal of statistical literature ● Study types and hierarchy of evidence ● Risk (relative risk, odds ratio, intention to treat, numbers needed to treat). Incidence and prevalence
Textbook review: Chapter 13. Relative Risk versus Absolute Risk
Class 12 Reading: Fineout-Overholt, E, & He, Z. (2017). Research 101 Forest plots. American Today, 17 (5). https://www.americannursetoday.com/rea search-101-forestplots/ Grant, M., & Logan, A. (2009). A typology of reviews: An analysis of 14 review types and associated methodologies. Health Information and Libraries Journal 26, 91108. https://onlinelibrary.wiley.com/doi/fu ll/10.1111/j.14711842.2009.00848.x Class Notes: Systematic review and meta-analysis ● Conducting a review ● Limitations and considerations ● Interpretation of meta-analysis results
Textbook review: