The study design

The study design

The study design that I will adopt is the survey design that involves sampling three out of the nine villages based on demographic and socio cultural factors and randomly selecting the households in each predetermined village to collect data from mothers of children below the age of 5. The questions I will ask in the questionnaires will be based on research variables to illicit response that can be analysed to measure the study objectives and test the null hypothesis.

3.3 Study Population and Sampling frame

The target population in this research will include women with children below the age of 5 in a selected household. The total number of households and sample population will be 14 and 909 respectively drawn from the three sub locations of Korogocho and one village in each sub-location. The households that will have women with more than one child below the age of 5 will be treated equally with the women with only one child even though diarrhea will be considered to be in either of the children. The intended sample size is 90 respondents that will be distributed across the three sub locations of Korogocho location.

3.4 Inclusion and Exclusion Criteria

Households included in this study must have members who have lived in Korogocho slum for at least one year and must be serviced with a source of water supply whether from tap or communal tap. Most importantly the household must have a child below the age of five years and the respondent (the mother) must be willing to participate.

3.5 Sampling Procedures and Sample Selection

Korogocho slum is purposively sampled and selected from other slums like Mathare, Kibera and other unmentioned slum within Nairobi County due to the fact that I had prior knowledge on the administrative and location of the slum better than other slums; this is attributed to the fact that I once resided in Korogocho slum. Three out of the nine villages of Korogocho will be selected on the basis of social and cultural attributes like ethnicity and mode of living.

The selection of respondents will be done through systematic random sampling. Every third household with a child under five years will be selected and in case there is none I’ll move to the next household with a mother of a child under five years until a sample size of 90 is achieved.

3.6 Sample Size Determination

For the results of the study to be used to make an accurate generalization at 95% confidence interval, the sample size will be found using Fisher et al (1998) formulae of calculating the wanted sample size as illustrated below based on the established 31% diarrhea prevalence among the under-fives in Korogocho according to APHRC 2002.

n=z2pq/d2

where n is the desired population sample size (when the population is greater than 100,000)

z= to the standard normal deviation (1.96) that corresponds to 95% confidence interval

p= the proportion target population that is estimated to potray a particular characteristic that of 31% of the population of children below the age of 5 with diarrhea infection (APHRC 2002)

Thus p= 0.31

q= 1.09-p= 0.69

d= which is the degree of accuracy desired that is 0.05.

The equation will give a sample size of 328, with an estimate of household ratio of to the total population of 1:3.5 (2/7), the household sample size will be attained by multiplying the population size by the ratio of household size to the total population as demonstrated in the equation below.

n=2/7(z2pq/d2)

In this case, a sample size of 93+ or – 3 will considered significant to give a fairly accurate results. The sample size will be distributed equally in the three sub-locations and the distribution will be based on the number of houses to the total sample size and the total number of household.

3.7 Data Collection instruments, Methods and sources

The primary data for this research will be collected through the administration of questionnaire to the respondents. A semi structured questionnaire will be orally administered to mothers with children below the age of 5 to ensure proper understanding of the questions thus giving responses that are not biased. The questionnaire will contain the questions linked to the variables of the study as stipulated in the study questions. The variables include; source of drinking water, level of the mother’s education, treatment of drinking water, access to sanitary facilities, hand washing, level of income, hand washing and hygiene amongst others. The questions will be formatted to collect data from both the environmental determinants of diarrhea and socio economic determinants. The secondary data will be sourced from books, articles, journals, academic thesis and even the hospital records.

I will also make observations of the community and household environment to reinforce data that will be collected through the questionnaires. Spot check observations will be conducted to support the data collected from the questionnaires

3.9.1 Household Data Collection Methods

Households will be visited to list and identify eligibility. After initial listing, two trained research assistants will then visit each household for data collection and sanitation inspections using the structured household questionnaire. However, before commencement of the data collection, respondents’ consent to participate will be soughed. The respondents will be asked to identify the source, frequency and quantity of water they use in the household as well as type and form of sanitary facilities available for their use. Sanitary inspections will be performed through visual assessment of the infrastructures and the sanitary state surrounding the household water supply, water holding containers; among other household sanitary conditions taking into account the sanitary aspects and practices in water handling that will be likely an actual or potential risk to water quality, health and wellbeing of the child. Most importantly, the respondents (mother) will be asked if any of the children in the household had experienced diarrhea in the last 6 months prior to the visit. If the answer was yes, the respondent was asked the age of the child at the time of infection, whether treatment was sought and from where and the symptoms and signs of the episode.

3.9.2 Hospital Data Collection Methods

Health facilities operating within the selected estates will be visited to review childhood diarrhea-related morbidity and mortality records. The selected health facilities were purposively and conveniently selected and matched with the particular selected estate as identified. Thereafter a structured data capture form was developed to collect the data. The key parameters for this review, will include the residence of the visiting child, diarrhea had been clinically diagnosed, age in months of the child suffering from diarrhea episode and the duration of diarrhea episode. This information will then be used to match the hospital case with the selected estate. All cases that will have no information on residence will be dropped from this study. Further, each case will be reviewed to identity the sex of the child, age in months, date of the onset, location and residence of family, duration of the episode and treatment outcome. The data were then cleaned and the information entered on to the research database.

Table 1.1: Health facilities that will be surveyed in the study area for review of the paediatric records.

Makwk Community Health Centre

Mwangaza Ulio Na Tumaini Health Centre

St. Mary’s Health Centre

EABL Health Centre

Provide International Health Centre

Korogocho Health Centre

Crescent Medical Aid Kenya Korogocho Clinic

4.0 Data processing and Analysis techniques

The primary data from the questionnaires administered to the mothers with children below the age of five will be cleaned of any errors and omissions and then coded for entry into the computer system. The available raw data that will be extracted from the questionnaires by entering the findings in an excel sheet after which the data will be organized, coded, filtered using the SPSS software. Both descriptive and inferential statistics techniques will be used to analyse the coded data. The data will first be analysed using descriptive statistics and cross tabulation. Which will give first-hand information about occurrence of diarrheal disease among children below the age of 5 for the respective characteristics. Descriptive statistics of percentages and frequency tables will be calculated for each variable in relation to the occurrence of diarrhea in children under five years. There will be further analysis of the data by use of chi square statistics techniques to give the relationship between the independent variables.

The qualitative variables like the occurrence of diarrhea in children below the age of five will be operationalised by setting frequencies of respondents who will report diarrheal cases and those who will report non-occurrence. Both the responses will be coded differently in the SPSS software to perform a quantitative analysis test using the chi square statistics. There after the data will be put in the contingency tables to obtain the expected values and the degrees of freedom. The Chi square analysis will use the formulae of observed and expected values to get the critical value (2x).

The chi square will be used in this study since it is easier to compute than other statistics and can be used with data that has been measured on normal scale. It also doesn’t make assumptions about the distributions of the population; other statistics assume certain behaviours on the distribution of the population’s normality.

REFERENCES