Epidemiology Assignment Solution
11 Pages3090 Words181 Views
Added on 2020-10-05
Epidemiology Assignment Solution
Added on 2020-10-05
ShareRelated Documents
Practical Epidemiology - SPSSIntroduction:Portfolio Assessment is a kind of assessment that have large number of meaning and itsprocess includes methods that can serve a variety of purpose with different result so that accuratevalues can be determined. In general it is stated that portfolio assessment could be examination ofstudent-selected illustration of activity experiences and written document related to result beingreceived by them. It is totally helpful and supportable in progress that further support in achievinggoals, including student affectivity. Portfolio assessments are used for large-scale classification andaccountability intention. The report cover, preparation of epidemiological data for analysis withtypes of variable, predicator variable, statistical methods to determine missing data, Frequencydistributions, central tendency, standard deviation, standard error, variance and confidence intervals.Apart this correlation and linear regression and differences between groups and logistic regression.Topic 1: Types of variables, recoding outcome and predictor variablesContinuous variable in to a binary variable: From the number of question it has beenstated that the survey question ‘If currently smoke, how many cigarettes per day?’ would be aresult in a continuous variable, according to Andrew Messing (2011) a continuous variable is said tobe one which have a continuum and which could have an infinite group of values so that betteroutcome can be extracted. In general a continuous variable is defined as a variable that have aninfinite number of possible outcome or values which is totally opposite of a discrete variable thatcould only consider only specific number of result. The above asked question aboutthe amount ofcigarettes the smoke per day? meets the criteria of continuous variable such as theoretically anindividual could have smoked one to the infinite amount a day. From the variable, a researchquestion that is needed for the number of cigarettes that are smoked a day to be reduced into othercategories which will result in binary variable can be “Do you currently smoke?” This is meantthat actual data is needed to be divided into two groups such as those individual which say that theydo smoke they would answer yes to the question and those individual who do not smoke wouldanswer to noIt is very important to transform the related rationale into other variable as the infinite set ofvalues would be reduced into different categories. In statistic a variable that are taken one of thelimited data and have a definite number of reliable outcome from each individual or group ofobservation that shows the result for a specific category depending on few qualitative property areknown as categorical variable. According to Maeve Duggan (2013) variable with more or tworesponses option are consider as binary variables.Continuous variable in to an ordinal variable:From the number of question it has beenstated that the survey question “If smoked in the past, how many in the past?” would be a resultin a continuous variable because the factors have an infinite number that can be ascertained as apossible outcome value. They are mainly the opposite of a discrete variable so that it can take morethan one value in the same series of inputs. So this survey question too meets the criteria ofcontinuous variable because theoretically an individual could have again smoked 1 to an infiniteamount a day. A research question that is needed would have to bethe split of the continuousvariable into categories that would make an ordinal variable would be “How many cigarettes doyou smoke a day?” because then there would be a ranking of groups which will result in binaryvariable for example 0 - don’t smoke, 1 - 10 cigarettes a day, 11 - 21 cigarettes a day, 22 - 31cigarettes a day.
Ordinal variable in to a binary variable: Ordinal data is statistical information in whichthe variable have the natural, ordered categories and the actual difference between two categories ismissing. The survey question “What type of diabetes do you have?” would be a result in a ordinalvariable because individuals can answer categories of I do not have diabetes, Type 1 diabetes andType 2 diabetes because it can be ranked from not having diabetes to what type of diabetes theyhave making it an ordinal variable. A research question that is needed to be reduced into othercategories which will result in binary variable can be “Do you have Diabetes?”this would dividethe answers into categories so then individuals can give the answer of “yes or no”.Ordinal data isalso is a statistical information in which the variable have the natural, ordered categories and theactual difference between two categories is missing.Nominal variable into a new nominal variable: A nominal variable can also be called acategorical variable and an independent variable they have two or more categories not having anykind of order they are variable that have a no numeric value to them Statistics How To (2019). Sothe survey question “what is your religion” would be a result in a nominal variable becauseindividuals can answer categories of which religion Christianity, Islam, Judaism, Sikhism etc. Aresearch question that is needed to be reduced into fewer categories that will result in new nominalvariable can be “Are you religious” this would divide the answers into categories so thenindividuals can give the answer of “yes or no”. Topic 2: Approaches for missing data and outliers
From the above table, it is has been seen that mean substitution the standard deviation hasbeen calculated. Both age and salary is minimum than it was earlier. This tends to make use ofstandard deviation refers to how spread out a data set is from the mean average. Also, the total,much data set deviates from the overall standard. Missing data and out liner's ends to create a wideroverall distribution for data sets. So, this is not always beneficial and can give an inaccuratedemonstration of the mentioned data series. Mean substitution tends to works by replacing overallout liners and missing values with the overall means on average values Allison, P. D. (2001). Smallstandard deviation means that the chances of risk is very much low in coming period. The tableshows total number of observation is 957 from which different kind of frequencies have beendetermined. Such as mean is equal to 53.37, the result for standard deviation is 38.226. Therefore,from the result, the missing value is 44 and the percent is approx. 4.4%. The total number ofobservation for children, salary and current smoker are 931, 941 and 495 respectively. With thesupport of frequency valuation the missing value of for all the factors have been determined.Missing value for salary is 70 that is equal to 7%, number of children is 60 that is equal to 6% andfor current smoker the value missing is 560, which is 56.5%. The missing data causes a loss of data in research outcome due to which the outcome areunrealistic and makes evaluation of data more difficult. The main advantages are that this issupportive in case if it recollects the maximum number of circumstances according to analysis, butmakes it unreliable to associate and compare examine of two data set as every time the sample isdifferent. Another way to compact with missing statistics is through regression founded singleaccusation to forecast the missing reply. This works well because it uses information from othersurvey questions to provide a guess to unanswered questions although, it cannot be fully accurate.
From the above graph the basic aspects which is seen in the above is relation which is relatedwith the salary or age outliner's which is basically arises because of an error in the per in the personfilling out the survey or by a person inputting the data into a software system such as SPSSincorrectly. Such instances of human error in data entry are common explanations for outliers. Forthe lower of the salary outliers, the person may have answered truthfully and the data may havebeen put in correctly they may just have a very high salary Hyun Kang (2013).Reason to replace the outliner's in the salary variable with a missing value because it is missing atrandom, which means that using other survey questions, an approximated age can be estimated.The higher salary outlier should be coded as a missing value because it is missing completely atrandom, and would be very difficult to approximate through other survey questions. The lower ofthe salary outliers should be left in because although it is an outlier, it is borderline and still aconceivable value.Topic 3: Descriptive statistics for nominal and scale variablesThesurvey question I had chosen valid weight (kg) in estimated> 130kg is distributed with a skewness of .360, it is easy to determine from the histogram that thegiven hill shaped curve that shows the values small start, step by step the result for the taken outputkeeps on increasing and then reduces towards the end of the set. It is observed that when the data isdistributed evenly, the mean and median measures of central tendency are relatively close together.Histograms are useful because they can be easier to read than tables, charts they also show moreinformation about data SPSS tutorials (2018). Histograms can show more info about data, we canestimate a variables standard deviation, mean and skewness. From the above table, the total numberof valid observation are 8285 and the missing value is 996. The value of calculated mean is 67.0182
End of preview
Want to access all the pages? Upload your documents or become a member.
Related Documents
Practical Epidemiology-SPSSlg...
|26
|2909
|30
Epidemiological Studies and Housing Improvementlg...
|8
|2635
|98
SPSS Assessmentlg...
|32
|6201
|31
Western Sydney University Statistical Methods in Epidemiology Assignment 2022lg...
|13
|3112
|11
Cross Sectional Survey Method Reportslg...
|5
|1042
|19
Logistic Regression Analysis for Survey Data Variableslg...
|8
|1152
|205