Dental health in Glasgow
Staff at Glasgow Dental Hospital carried out a study to examine the dental health of 5 year old children living in Glasgow. Each of 360 children had a dental examination under standard conditions and the number of decayed, missing and filled teeth was recorded. Also recorded were the deprivation category of the area (postcode sector, such as G12 8, or KA25 7) in which the child lived, and whether the child was currently registered with a dentist.
You have been given a subset of 100 of these children in the dataset with matriculation number in it. The datasets are in tab-delimited text format (“.txt”) – the file you need has your matriculation number on it, for example, if you matriculation number was 123456789 then your file would be called “AS_123456789.txt”. Also save the file called Assign1.tpf as well (it can help when loading the data into SPSS – see below).
To open your dataset in Minitab, go to File > Open Worksheet in Minitab, choose filetype of Text (*.txt), select your dataset and click Open.
To open your dataset in SPSS you have two choices:
1. Open the data in Excel and in Excel save it as an Excel file. Then open the Excel file in SPSS the usual way. You could also do this via Minitab.
2. Go to File > Read Text Data in SPSS, choose your file and click OPEN. This will open a window called “Text Import Wizard – Step 1 of 6”. Click the Yes button to the question “Does your text file match a predefined format?” and then click on Browse and choose Assign1.tpf and click Open. Then click Finish twice and the data should open. Once you have done this it makes sense to annotate the data (e.g. indicate that 1 means boy and 2 means girl for Sex) and then save the data with an appropriate name in the .sav format before you do any analysis. To open the data in Minitab, having altered the dataset in SPSS, just copy and paste the data directly from the SPSS data editor into Minitab (you could also save a copy of it as an Excel file, and then open that in Minitab).
Research question of interest:
• Do 5 year old children living in deprived areas have worse oral health than 5 year old children living in more affluent areas?
Variables in your excel file:
RefNumber Child reference number (for admin purposes)
DepCat Deprivation category (on a 1-7 scale, where 1 is most affluent
and 7 is most deprived). In the catchment area for this study,
all the postcode sectors were in DepCat 4, 6 or 7
RegCat String (character) variable, with 3 categories:
NotReg – never registered with a dentist;
Lapsed – previously registered but now lapsed;
Reg – currently registered with a dentist
Sex 1 = boys, 2 = girls
DFMT Number of teeth with active decay, that are filled or are missing
pH pH value of saliva in mouth – the lower it is, the more acidic.
Questions
NOTE: for Q2 and Q4 you are expected to follow and report on each stage of the Data Analysis Algorithm shown on p1 of Workbook 3, and ensure you cover both confidence interval and hypothesis testing approaches.
1) Using either Minitab or SPSS, obtain appropriate descriptive statistics for the variables Sex, DepCat, RegCat and DFMT. Provide a short interpretation of the output you produce.
[6 marks]
2) Using Minitab, carry out a test of 2 proportions using a significance level of 5% (and produce the 95% CI) to investigate whether those living in the most deprived areas (DepCat 7) have different rates of good oral health (having no decayed, filled or missing teeth) compared to those in less deprived areas.
[10 marks]
HINTS:
1) Create a new variable (GoodOH) that takes the value 1 if DMFT=0 and the value 0 if DMFT>0. The categories of this new binary variable can be thought of as “excellent oral health” and “not excellent oral health”. You can use Data > Code > Numeric to Numeric to create the variable in Minitab and Transform > Recode into Different Variables to create the variable in SPSS.
2) Create another new variable (Poorest) that takes the value 1 if DepCat is 7 and 0 if it is less than 7.
3) Note that you can use summarised data in Minitab to do the test of 2-proportions, so if you have been using SPSS up to this point you do not actually need to copy the data across to Minitab.
3) How else could any relationship between good oral health (as measured in GoodOH) and deprivation (as measured in DepCat) be investigated?
[4 marks]
4) Using a 2-sample t-test and confidence interval in SPSS, investigate whether there is a difference between oral health, as measured by DFMT, in the poorest areas (DepCat 7) and in less deprived areas.
[10 marks]
5) Is the mean a good measure of central tendency for the DMFT index? Explain your answer.
[2 marks]
ORDER THIS ESSAY HERE NOW AND GET A DISCOUNT !!!