Home
This site is intended for healthcare professionals
Advertisement

Module Session 2: Multiple Statistical Tests

Share
Advertisement
Advertisement
 
 
 

Summary

This on-demand teaching session is for medical professionals who want to explore in-depth the concepts of descriptive and inferential statistics, which are essential tools in medical research and practice. The session elaborates on different types of variables, such as numeric and categorical, and their variables. The session also focuses on techniques for analyzing, interpreting, and presenting data, like measures of central tendency, standard deviation, frequency tables, and the like. Attendees will also learn about different statistical tests used for assessing associations between variables including chi-squared test, T-tests, and ANOVA. This session is ideal for professionals who deal with data regularly and are looking to strengthen their statistical analysis skills.

Generated by MedBot

Learning objectives

  1. To understand and distinguish between different types of variables such as numeric variables, categoric variables and nominal variables.
  2. To learn how to transform and code variables effectively for variant analysis.
  3. To comprehend descriptive statistics including measures of central tendency, measures of dispersion and descriptive statistics as applied to both categoric and numeric variables.
  4. To learn how to conduct and interpret frequency analyses.
  5. To understand the principles of inferential statistics, including linear regression and chi-square tests, and to effectively apply these principles to real medical data.
Generated by MedBot

Similar communities

View all

Similar events and on demand videos

Advertisement
 
 
 
                
                

Computer generated transcript

Warning!
The following transcript was generated automatically from the content and has not been checked or corrected manually.

Six and 34. Uh No. Is a doctor studying a dog. Unfortunately, he might had with him uh uh variable types, two types of variables, categoric variables, numeric variables, categoric variables, no, no variables and so on. However, education level, education level, masters, masters, bachelor's, bachelor's. OK. Uh The number and a number of people who will discrete number of correct answers, the numeric variables uh presentation, numeric variables with variables, presentation continue with it. Continue with A N medical and same uh explanation. OK. Well, a lot touch your tongue in no. OK. OK. OK. OK. The variables. OK. OK. So numeric data continuous with the OK. Analysis analysis. OK. Nominal variables. So type 10 and the nominal variables. Uh 22 categories, the di di variables. A yes or no, a gender, male, female. One, yes or no. OK. Hello, education uh did not graduate from high school, high school, graduate some college or post high school education and college graduate categories, college graduate, wha some college plate high school or did not graduate from high school. Hello en in order. OK. A stable, a very bad certain scales, your status of satisfaction. I was very satisfied with all. Not satisfied. The ma not satisfied. Very satisfied, satisfied, not satisfied. OK. OK. Know how to on uh on, on is a uh Yes, yes, I should. OK. Tell me um just wait a minute. OK. Transform, transform one with female two. OK. Little hand foot and a trans they code into different variables. OK. Transform into different question mark analysis in gender name, gender, new label, gender but help to change gender old and new values. But uh a excel the arm copy LMA questionnaire. OK. OK. OK. To one female female Google form C A female E two analysis. OK. To continue. OK. Um OK. Ha 3 million orange way. OK. One mail. So with two who were female we add then we click. OK. Is a female? This this view value labels. OK. Male or female. And I'll take simple example is down to three or four or five at dot OK. Hey. OK. OK. Uh um OK. The script of statistics like categoric variables uh statistics, statistics, variables. OK. Uh frequencies, percentage or proportion A A frequency analyze the script, statistics, frequency. OK. Analyze OK. Transform. Analyze analyze descriptive statistics, frequencies, descriptive category variables, descriptive linear percentage with frequency. Hello quent analyze descriptive statistics, frequencies. Hello ma ma ele. OK. OK. Display frequency tables. OK. OK. Out. OK. Uh that the frequency he will percentage percentage minute total sample size. OK. Type A is the interpretation interpretation. It's the interpretation 29 out of 67 females must not in no 17 people ma OK. Uh So 12 view. OK. Yes. OK. Tell me. Hello. Descriptive statistics continue with our numeric variables. OK. And frequency continuous variables or numeric variables. The statistics will will. OK. OK. OK. So OK. Um OK. Reporting plus or a standard deviation or minus range. That's the same thing for the media. OK. Uh Measures of central tendency measures of dispersion. Uh OK. Uh variable. OK. I Mr He will continue as variable. OK. Who had a example in me? OK. Who were analyzed the script statistics frequencies kin will median will mode analys the script of statistics, frequencies. OK. Hey BM I but ambiguous had a statistics but OK. The statistics but measures of central tendency, OK. Discontinue but OK and so on and so forth. Um actor, will he shame the actor? Will he explode? Um My fruit, standard deviation I QR variants and so on. OK. Uh analysis anyways have helped analyze descriptive statistics explore OK. He he me uh uh little heck after OK. We R BM I of how to build dependent list. OK. How to build dependent list but then how OK. En des des OK. Standard deviation, minimum maximum, minimum 18.3 34.7. OK. Range maximum, not a minimum will enter IQ RM. OK. Uh extended deviation um visual uh statistics. I see you're welcome the frequency percentage or proportion. OK. Uh standard deviation with uh analyze the script statistics, frequency. OK. You can ask me OK. Illustrations plus or ill. I keep the graphical illustrations like I had what uh charts uh no percentage continuous data to report them as histograms. OK. Type. Um OK. OK. And analyze the script statistics, frequencies, frequencies, percentage of frequency category variables who who analyze the script statistics frequencies. OK. The $4 gender. New Llao. OK. Uh OK. Thank uh and no display frequency tables contract and to my yes, Mike. Oh OK. Tell me uh display frequency tables checked the OK. Uh S OK. Obama Obama control. OK. OK. A a no, no A OK. Type the Humulon select with a food insert, the lax les frequency G OK. The gender distribution, the number of participants based on gender. OK. Uh chart a percentage. OK. No. Um uh it's the same um missing values and 25% missing. OK. Value morning to nine females. OK. Would I hike to percentage? OK. Ella select po insert two other colors. For example, uh you know, it would be better to be uh easier for the for the person. It's very fine uh word document. OK. So I guess it's very easy. Um uh Us Legacy dialogues Histogram. OK. Graphs Legacy dialogues Histogram for or this plain or milk, milk. OK. Distributed along the frequency uh easier for you to report. OK. Uh Hello, I will leave you with my colleague Daniel. Influential statistics. Thank you, Rosanne. So much. You're welcome. It had a so I'll uh you can ask. OK. OK. Hello everyone. Bye. And just like yes. OK. So descriptive statistics usually test is a uh association by two variables or is a factor outcome. So, well about the introduction is really uh linear regression. So the flow chart type of test the uh reference uh dependent variable outcome, disease, independent variable exposure. So methyl, the se and no continuous to category variable. So outcome continuous is I think an example who uh is a measure like blood uh blood pressure become continuous. So uh she uh dependent variable uh category two groups or two groups. Uh category and gender two groups, female females, two groups who and continuous. So if I'm not on the flow chart should be our test. The outcome category well exposure, categoric the test. A square test. So OK. Um Bill uh Bill for so C square test two way uh and is a association by two category variables. Um disease. Yes to no gender, male, female. So they had high uh square test. OK. The conditions who we know uh no. Category one down I, yes doctor. She is not. Oh OK. Yeah. That way you uh have a nice OK. But um and C two groups, three groups. So ana uh um uh data OK. It's my inferential statistics. So um depressive symptoms. Yes. Yes. OK. So uh uh depressive symptoms is a outcome. Social media use um depressive symptoms. OK. Depressive symptoms and four groups, no symptoms, mild to moderate to severe who social media use uh low users, high users, average users who are very high users. So test was analyzed oh statistics who um was cross step outcome and uh exposure or independent variables. Only a hard depressive uh symptoms, scale outcome and how to build colons II been uh and uh who social media use? Ma uh statistics. OK. When square test, OK. Percentage social media use uh stage and I depressive symptoms. OK. When hot con continue while in hot. OK. So Bill output he long has three tables. The first table, the frequency cases, missing cases, cases and zero cases, participant questionnaire. A tiny who we cross tabulation, who we few descriptive statistics, frequency who he percentage depressive symptoms. 95.5% of uh low users had no symptoms, no depressive symptoms. We must have done um that high users ma and 53.8% of high social media users had moderate depressive symptoms. So had a pure statistic uh description reporting but she is a association by the by the depressive symptoms. Will social media use uh square test it uh table. Uh This test is ap association with a small P value and he probability that 01. OK. High P value is a association with uh I had AP point OK. OK. The lower the P value, the higher the significance association with uh uh lower than 5% of 0.05 significant difference association. So significance significance value. OK. Honey make sure it's equal to 0.000 and 0.00015%. Man test is significant significant difference. OK. The percentage depressive symptoms we kill category and social media use um uh the prediction and the association be how they do V Jon. Yeah, you had a no, it's it. OK. So here the whole will uh square how many of the steps get? One is a interpretation. Hanna Hanna in uh flower, two independent sample T test outcome, continuous exposure category of two groups, two groups, gender, male, female, how the two groups, so show how many uh two independent sample. The US uh who we, we can handle dependent variable will be continuous uh in the blood sugar level measurement, our BP, our scale down total score. OK. Me and we can gender and independent variable. OK. Males will mean females. OK? Is a uh difference will mean a difference significant he is significance. So the top the S PSS OK. I come in the steps, get home, come in uh outcome, memory satisfaction, the memory satisfaction who we taught at men. Um uh men uh come scale memory satisfaction scale. OK. A participant can for OK. L is a heal mean it right here. 100 males, 100 females analysis now compared compared means who had an independent sample. Variable outcome. OK. Who have memory satisfaction? No transfer who like gender who will independent variable uh no transfer. Grouping, variable two groups OK. Group one, group two, group one in male with group two in females. OK. OK. One or two equal female. So was defined group group one equal one male, group two equal to females. Men continue over. OK. OK. Uh tables first table who come in description the frequency and male frequency and female mean standard deviation. So uh comparison by how 246 to 42 is a significant difference. A few significant difference by me. OK. Table, I assume that two parts test T test for equality of means. And um 5% is a significant element. 0.05 or 5%. OK. Bottom line. OK. Significant significant difference with VV assumed equal vance assumed as a significant difference. Equal var is not assumed. So bottom line 5% no top line. So HP value P value equal 0.0 29 5%. There is a statistically significant difference between the mean memory satisfaction. Male who handle female. Uh memory satisfaction is higher. Handle male is significantly higher than male. OK? Can wait any test and uh independent test you had. Thank you so much. OK. So how can get one until the uh kill the steps with the interpretation? Hello uh I uh one way and one way a objective independent sample T test. OK. There's a difference in two independent sample. T two groups, category one and two and three groups with. So who and uh uh comparing means between three or more independent groups. OK. Dependent variable or continuous uh variable outcome. Memory satisfaction continue is variable memory satisfaction. It right here significantly between a different stages, depressive symptoms. OK. For an example, depressive symptoms will be category variable only for um or groups. One nick was one way and no bit dependent transfer. Uh memory satisfaction, continuous variable outcome. Who will factor independent variable or will uh exposure? OK. Um Options. OK. Descriptive woman hot. OK. One way and over uh three tables. OK. High frequency will mean with standard deviation 95% confidence in two minimum max. So groups. OK. I can, I can value in 0.05 and at least two groups. So give results, statistics interpretation of homogeneity of variance might be difference with variances between any groups. So significant element uh 5%. So is based on the P value equal 0.7 21 5% PP value 0.0001 highly significant. OK. Uh Highly significant difference will mean between at least two groups, 234, at least two groups means the Anova Metformin Anova I said we'll see, OK, you only come on the steps and uh the interpretation at three tables. Pearson correlation um flow, the correlation outcome continuous will exposure continuous. So uh Pearson correlation or correlation. He here the relationship between two continuous variables. OK. Positive correlation, negative correlation, positive correlation, correlated decreases negative correlation. You know that's not so come in uh in um SP PPP value significantly 100 and 5% correlation correlation coefficient correlation association. So uh hello, how did the steps of S PSS? Um memory satisfaction. OK. Well, somatic symptom. OK. Hold it nine is the association or correlation analyze, correlate women did name memory satisfaction will somatic symptoms. Uh and Peterson. OK. Um OK. Mister Bye. OK. So uh two box correlation coefficient. OK. Bad then. Um my heart. OK. Hi, I had been out good. OK. So I will share now now OK by your memory satisfaction with somatic symptoms can hide them and truthful. I hide them. That's a how sick IP value P value is 0.000 and a high significant correlation by somatic symptoms. Well, memory satisfaction. OK. Hi. Um color point. OK. Like an LP value is a correlation first and correlation and correlation coffi equal minus 0.3 321 correlation always shortly. The coefficient is negative, negative correlation, memory satisfaction and somatic symptoms and which is common logic. So interpretation interpretation uh minus 0.3 21/10 weak correlation. So correlation correlation is weak uh interpretation correlation. Uh no uh Rola linear des uh linear regression. Who we met a computational model and a linear relationship between two continuous variables. Memory satisfaction with somatic uh continuous. OK. Um So exposure prediction, memory satisfaction, you know, somatic symptoms, somatic symptoms, uh um prediction value about memory satisfaction, sores and no um G equation. OK. Equation nine. So the simple linear regression prediction and value of outcome ta ta and exposure of the predictor line. So a condition that is dependent, variable, continuous, well independent, variable, continuous. So how many steps, how do, how do S PSS who have enjoyed the somatic symptoms of low prediction and memory satisfaction? Analyze? OK. Very dependent when dependent um uh value story, memory satisfaction, somatic and in independent. OK. Hot. OK. I had four tables table. OK. Our variables entered or removed, dependent variable, memory satisfaction, uh somatic symptoms. Hello, tell you what had the mal summary. OK. RR with somatic symptoms, variability value value. So in somatic symptoms and the so are square uh percentage. OK? And 100 0.103 and 10.3% humana and no systematic symptoms. They explain 10 10.3% of the variability of this outcome. So the uh reports come in a normal uh reporting square. How about that meeting? So meeting is been the association between two continuous variables. OK. Heart rate like memory satisfaction, outcome, well, independent variable symptoms and by uh equation uh no prediction and value outcome, basal and value an independent variable. So uh more summary, RR square with an independent variable and variability outcome. So uh keep reporting when all R square equal 10.3% and no somatic symptoms. OK. Explains 10.3% of the variability of the memory satisfaction score. Thank you. Hello is a heel model over the prediction. OK. OK. IP value AP value is significant. Metal model is a good fit to predict uh out equation to predict the outcome. So honey it's highly significant uh is a mechanic, significant regression line association between and two variables and um power point OK. Coefficient table. OK. I have a coefficient equation. Oh Equation is an independent variable. Uh Is it no significant significant man. Somatic prediction equation? The equation here, Y equal B zero plus b1 times X and Y who will outcome? Who will independent variable and Y dependent will independent. OK. So why we had example who memory satisfaction? The B zero constant co value. So by the little um uh 48.23 cof and somatic score minus 1.2 OK at times in somatic symptom. So somatic symptom OK. Is it been a no prediction that value about memory satisfaction? OK. Linear simple linear regression. Hello uh Metformin. OK. OK. So is small simple linear regression. Abnormal prediction outcome valuable outcome based only 11 value uh sorry, one variable, one independent variable, multiple linear regression continuous in um the prediction equation. So multiple linear regression that can, it allows to predict a dependent variable of outcome based on multiple independent variables. So he step by line dere but independent variables, the conditions, he achieve an outcome but he continuous but still independent variables. How name continuous categoric continuous variables as independent variables, continuous. OK. Age uh somatic symptoms, linear des equation. And it is a significant uh element 5%. OK. Is a significant multiple linear de So uh the example I had the SPS on family. So the interpretation OK. Med S PSS came, Nick was analyzed digression of linear uh line regression by memory satisfaction will age OK. Will age significant the prediction and memory satisfaction with multiple linear regression dependence. OK. So um multiple linear regression, multiple linear regression. OK. Uh Statistics significance. Oh OK. We had your heart. OK. So four table table, the dependent variable 10 table model summary commandment R square square 19.6% and how the uh the predictors. OK. And combined they explained 19.6% of the variability of the variation will score memory satisfaction. OK. Memory um model summary I know um interpretation is a model higher regression model. Is it a good fit to predict the outcome on the P value? Can it's highly significant and it it can, it can predict and coefficient equation the power point OK. Coefficient table equation value. OK. So somatic can uh significant but significant variables. OK. So, but a significant moment equation, the gender health evaluation with depressive symptoms, 0.05 significant equation governor with social media use 11 5% uh uh uh to generate the regression equation, line regression. And Y who will outcome with and, and B zero A kid who will constant will XX two, X three and independent variables. Mhm So the example B will be value with serial equation rather than Y memory satisfaction. Men. Constant. OK. Who men? Uh gender health evaluation when depressive symptoms might coefficient about oh in a new equation. OK. I know that it had, I do that the ACL uh test. Thank you. My uh my leg had come in and I give some reports. Uh OK. It's called oh but uh uh for recording and it had the G OK. Uh uh uh uh I wear she um it's a structured document. Document results will find organ, organ parts, uh interpretation, introduction and a conclusion is used and the recommendation decision making uh statistical results. Uh um The report of the research article is uh other than uh part of our uh of the uh of the topic will objective. OK. So on the top uh uh statistics account system objective uh in case of accountability report aimed to measure social accountability, faculty, university and all the students with disease or uh hi there. I introduction methodology, methodology here and materials and methods and learning environment questionnaire. We want to measure learning environments but Faculty of Medicine and we use the questionnaire calculation uh group or approval target population. No ie learning environment, learning environment. Variables, outcome category continuous variables, you can the results to the main variables or interested variable report the outcomes, disease process and learning environment. Uh In case conclusion, the limitation results in numeric results. Graph chart and table will will figure figure graphs and charts will highly recommend that I mean to figure 1212 in two A two B by histogram uh uh very simple and interpretation 0.0 point. Um in case of scales, the total score means a male individual score. And um mhm the first to the head of my door. Uh so no. Oh do do uh uh uh in case of the script of statistics, C variable frequency variable reporting has standard deviation. A statistical test reporting that means standard deviation uh uh and regression um uh reporting the BCI value adjust R the example C variables reporting ceg frequency that each group size to con uh conclusion here. General conclusion of conclusion. Oh good. Oh Not good. Oh Excellent. OK. Um Not this way. No strength of the results, limitation of the results, strength of the results uh uh processing target population questionnaire recommendation, Faculty of Medical Science. Uh He reporting a lead tables and for example, had a social accountability. Um How, how long ago, OK. How she share screen? Yeah. OK. Social accountability questionnaire report and so social uh of medical school assessment materials and methods. OK. Uh Students for the Faculty of Medicine and seek to inspire and inform educators uh healthcare leaders, objective, educator, healthcare leaders uh performance, uh excellent assessment to the social uh agree, disagree et cetera or no, somewhat good and excellent from no to uh excellent uh a total score. OK. A total score average then uh then minimum and then maximum um in it group interpretation, interpretation total a total score, social accountability I eligible missing values. Uh uh A kid who represented my uh standard deviation uh embellishment results show that participants uh in the range one in range two. OK. Uh uh On reporting descriptive statistics like continuous body is an item question will mean understand the deviation with the results of the students distribution is scored are represented in figures too. Participants range categories, categories, uh frequencies will percentage come in of students, individual elements idea the individual the questions is for along uh the percentage of the questions. One somewhat good and excellent results in the question one question, the students are asked their institution percentage good or excellent. Uh clear social mission. Uh they, they can and clear social mission report. Uh uh School of Medicine belongs to the third group conclusion and interpretation, interpretation and doing well and social account uh with the strength and uh um conclusion limitation, a recommendation, the recommendation to implement mechanisms, recommendation conclusion in in the social account uh learning environment, Faculty of Medical Science, the target population and go on groups uh 1231234 overall interpretation questionnaire. Do each results questionnaire uh subgroups do uh uh questionnaire questionnaire, uh a report and a can a can handle the highest score domain. OK. Can have a positive perception of learning and teaching subgroups as per OK. Uh the domain of the perception. Third year students. Yeah. But man, but she report no upper significance, statistical significance uh materials and methods. Um again and uh uh you can come in here. Thank you. Thank you. You're welcome. The session can add hardware. Very clear. Simple. OK. Do I see?