Analyze the results

To analyze a result one must look for

1. relationships

• directly related (positively correlated)
• inversely related (negatively correlated)

2. differences

3. Similarities observed from data.

Every Analysis must be accompanied with table or graph and most especially with statistical computation.

Example:

Increased Host Species Diversity and Decreased Prevalence of Sin Nombre Virus

Laurie J. Dizney and Luis A. Ruedas
Author affiliation: Portland State University, Portland, Oregon, USA
Vol. 15, No. 7 • July 2009

Figure. Results of the nonlinear regression analysis between species diversity (expressed as Simpson diversity index, D_s) and Sin Nombre virus prevalence among deer mice (Peromyscus maniculatus) at each of 5 parks in Portland, Oregon, USA. The best fit model was of the form Y = x / (ax + b), R² of 0.9994, p = 0.00001. The figure represents a summary of the results in that it shows the averages of all the seasons, in all years, in each park (indicated by circles). A regression using individual seasons and parks shows the same results.

2. Title:The effect of methylphenidate on oppositional defiant disorder comorbid with attention deficit/hyperactivity disorder

Analysis:

Twelve patients were screened, assigned to the treatment group and initiated therapy.

Two patients (16%) dropped out of the study because of non-compliance with repeated medical appointments. The sample comprised 10 patients, six male patients, four female patients, aged 6 to 14 years. The mean time lag between first and last evaluation was 113.7 days. The mean daily dosage of methylphenidate was 23.5 mg.

Nine of the 10 patients who completed the protocol stopped fulfilling diagnostic criteria for ODD in their post-treatment evaluation a month or more after their ADHD symptoms were controlled. Table 1 shows the ADHD symptoms before and after treatment.

Patients fulfilled an average of 4.9 criteria in their pre-treatment evaluation, compared to 1.8 criteria in their post-treatment evaluation. The difference between these averages, reached statistical significance (p=0.004). Since one of the patients was undergoing psychotherapy, we conducted the same analysis without this patient and it still reached statistical significance (p= 0.008). Table 2 shows the fulfillment of diagnostic criteria in baseline and end evaluations.