How to Investigate the Use of Medicines by Consumers
(2004; 98 pages) Ver el documento en el formato PDF
Índice de contenido
Ver el documentoAcknowledgements
Ver el documentoPreface
Abrir esta carpeta y ver su contenido1. Why study medicines use by consumers
Abrir esta carpeta y ver su contenido2. What influences medicines use by consumers
Abrir esta carpeta y ver su contenido3. How to study medicines use in communities
Abrir esta carpeta y ver su contenido4. Prioritizing and analysing community medicines use problems
Abrir esta carpeta y ver su contenido5. Sampling
Cerrar esta carpeta6. Data analysis
Ver el documento6.1 Introduction
Ver el documento6.2 Sorting and ordering data
Ver el documento6.3 Making quality control checks
Ver el documento6.4 Processing qualitative data
Ver el documento6.5 Analysing qualitative data
Ver el documento6.6 Processing quantitative data
Ver el documento6.7 Analysing quantitative data
Ver el documento6.8 Conclusion
Abrir esta carpeta y ver su contenido7. Monitoring and evaluating rational medicines use interventions in the community
Ver el documentoBack cover
 

6.7 Analysing quantitative data

In the manual analysis of quantitative data, it is best to first summarize the data in a so-called data master sheet to facilitate analysis. On a master sheet, all the answers of individual respondents are tallied by hand.

Examples of data master sheets that you can use to tally results from the illness-recalls used in your field work exercise are given on the following pages.

Data are easier to tally from the master sheets than from the original questionnaires. Straight counts can be performed for key variables, such as “health worker advice”, or “use of modern medicine”. Using the data in the illness master sheet you can calculate the following four therapy-choice measures:

1. Percentage of total episodes which are not treated. This is calculated by dividing the number of episodes which are not treated with any form of medication (including home remedies) by the total number of episodes, multiplied by 100.

2. Percentage of total episodes treated with traditional treatment. This is calculated by dividing the number of episodes which are treated with traditional treatment by the total number of episodes, multiplied by 100.

3. Percentage of episodes seen by health worker. This is calculated by dividing the number of episodes in which health worker advice is sought by the total number of episodes, multiplied by 100.

4. Percentage of episodes self-medicated with medicines. This is calculated by dividing the number of episodes in which medicine is given without health worker advice by the total number of episodes, multiplied by 100.


Questionnaire data may be compiled or tallied by hand instead of using master sheets if it is difficult or impossible to put the information (such as answers to open-ended questions) in a master sheet. Hand compilation is also necessary if you want to go back to the raw data to make additional tabulations in which different variables are related to each other. In a survey, it is often useful to have several master sheets, depending on the nature and objectives of the study and whether you want to compare two or more groups.

Note: Take great care when filling in master sheets. You should verify that all totals correspond to the total number of study units (respondents). If not, all subsequent analytical work will be based on erroneous figures. There should be special columns for ‘no response’ or missing data, to arrive at accurate total figures. If a research assistant is entering the data, you should randomly check 5-10% of the entries. You can also have the data entry done twice, by different assistants. By comparing the master sheets for inconsistencies you can eliminate errors in entry.

From the data master sheets, simple tables can also be made with frequency counts for each variable. A frequency count is an enumeration of how often a certain measurement or a certain answer to a specific question occurs.

For example:

Fever episodes treated with health worker advice

63

Fever episodes treated without health worker advice

74

Total

137

ILLNESS MASTER SHEET

Sheet number:

Date:

Respondent number

Illness episode, give a description

Treatment Yes/No

Health worker advice Yes/No

Modern medicine used Yes/No

Traditional medicine used Yes/No

           
           
           
           
           
           
           
           
           
           
           
           
           
           
           
           
           
           
           
           
           
           
           
           
           
           
           
           
           
           
           

MEDICINE MASTER SHEET

Sheet number:

Date:

Fill in one row for each modern medicine recorded.

Respondent number

Medicine name

Generic content

Illness for which it is used

Source

         
         
         
         
         
         
         
         
         
         
         
         
         
         
         
         
         
         
         
         
         
         
         
         
         
         
         
         
         
         
         

If numbers are large enough it is better to calculate the frequency distribution in percentages (relative frequency). This makes it easier to compare groups than when only absolute numbers are given. In other words, percentages standardize the data. A percentage is the number of units in the sample with a certain characteristic, divided by the total number of units in the sample and multiplied by 100.

In the example above, the calculation of the percentage answers the question: If I had asked 100 people who had a fever episode, how many would have answered ‘yes’? The percentage of people answering ‘yes’ would be:

A frequency table such as the following could then be presented:

Table 6. The extent to which health worker advice is sought in fever episodes (N= 140)

CATEGORY

FREQUENCY*

RELATIVE FREQUENCY

Health worker advice sought

63

46%

Health worker advice not sought

74

54%

Total

137

100%

 

* missing values 3


Note: Sometimes data are missing due to non-response or (in oral interviews) non-recording by the interviewer. Usually you do not use missing data in the calculation of percentages. However, the number of missing data is a useful indication of the quality of your data collection and, therefore, this number should be mentioned, see table above.

Be careful: ‘Don’t know’ is not to be taken as a non-response. If applicable, a category ‘don’t know’ should appear in the data master sheet and in the frequency table.

Cross tabulations

In addition to making frequency counts for one variable at a time, it may be useful to combine information on two or more variables to describe the problem or to arrive at possible explanations for it; or simply to compare between groups. For this purpose it is necessary to design cross-tabulations. To visualize how the data can be organized and summarized, it is useful at this stage to construct so-called dummy cross-tabulations.

A dummy table contains all elements of a real table, except that the cells are still empty. In the personal inventory of medicines, one of your objectives may be to compare the number of medicines that women have in their bags with those of men. A dummy table for this comparison is given below.

In a research proposal, dummy tables should be prepared to show the major relationships between variables.

Note: It is extremely important to determine before you start collecting the data what tables you will need to assist you in looking for possible explanations of the problem you have identified. This will prevent you from collecting too little or too much data in the field. It will also save you much time at the data processing stage. Take care not to embark on an unstructured comparison of all possible variables. The dummy tables to be prepared follow from the specific objectives of the study.

Table 7. Number of personal medicines carried in their bags to the course by men and women

 

NUMBER CARRYING LESS THAN TWO DIFFERENT KINDS OF MEDICINES

NUMBER CARRYING TWO OR MORE DIFFERENT KINDS OF MEDICINES

TOTAL

Men

     

Women

     

Total

     

When preparing the dummy tables, consider the following rules:

If a dependent and an independent variable1 are cross-tabulated, the independent variable is usually placed vertically (at the left side of the table in a column) and the dependent variable horizontally along the top of the table. All tables should have a clear title and clear headings for all rows and columns.

1 The dependent variable is the variable that is under study. The researcher does not control this variable but observes it. ‘Drug use in the treatment of fever cases’ is, for example, a dependent variable. Researchers are usually interested in the effects of other, ‘independent’, variables on this variable, for example, the effect of the educational status of the mother on drug use practices.


All tables should have a separate row and a separate column for totals to enable you to check if your totals are the same for all variables and to make further analysis easier. All tables related to each objective should be numbered and kept together so the work can be easily organized, and the writing of the final report will be simplified.

To further analyse and interpret the data, certain calculations or statistical procedures must usually be completed. Especially in large cross-sectional surveys and in comparative studies, statistical procedures are necessary if the data are to be adequately summarized and interpreted. When conducting such studies it is therefore advisable to consult a statistician from the start, in order that:

• correct sampling methods are used and an appropriate sample size is selected
• decisions on coding are made that will facilitate data processing and analysis and
• a clear understanding is reached concerning plans for data processing, analysis and interpretation, including agreement concerning which variables need simple frequency counts and which ones need to be cross-tabulated.


Data processing: manually or by computer

As you begin planning for data processing, you must decide whether to process and analyse the data: manually, using data master sheets or manual compilation of the questionnaires; or by computer, for example, using a microcomputer and existing software or self-written programmes for data analysis. Keep in mind that the people responsible for computer analysis should be consulted very early in the study, i.e. as soon as the questionnaire and dummy tables are finalized. Hand compilation is used when the sample size is small.

Before you decide to use a computer, you have to be sure that it will save you time or that the quality of the analysis will benefit from it. Note that feeding the data into a computer costs time and money. The computer should not be used if your sample is small and the number of variables large. The larger the sample, the more beneficial the use of a computer will be. Also be sure that the necessary equipment is available, as well as the necessary expertise.

A number of computer programmes are available on the market that can be used to process and analyse research data. The most widely used programmes are: Excel, a spreadsheet programme; Access, a data management programme; Epi Info version 6.04, a very consumer-friendly programme for data entry and analysis, which also has a word processing function for creating questionnaires, developed by the Centers for Disease Control, Atlanta, and WHO; and SPSS, which is a quite advanced Statistical Package for Social Sciences (by SPSS Inc.).

If you intend to use a computer, ask advice from an experienced person concerning which programme is the most appropriate for your type of data. Note that Epi Info may be freely used and copied. All the other programmes have copyrights.

Ir a la sección anterior
Ir a la siguiente sección
 
 
El Portal de Información - Medicamentos Esenciales y Productos de Salud de la OMS fue diseñado y es mantenido por la ONG Human Info. Última actualización: le 1 diciembre 2019