Pricing Strategy

Cost of analysis will be stratified into two categories:First category will include analysis results without any accompanying explanation or narrative of output tables; & second category will have succinct and relevant explanation/narrative for relevant points in output of analysis. Aforementioned categories are explained below:
rtarrow Hypothetical example: A study was carried out in 40 women to find out their hemoglobin (Hb) values. Researcher is keen to know descriptive estimates of Hb variable in total cohort of 40 women.
Results without narrative:

Statistics Std. Error
Hemoglobin level Mean 12.24 .92
95% confidence Lower Bound 10.37
Interval for Mean Upper Bound 14.10
5% Trimmed Mean 11.76
Median 11.55
Variance 34.11
Std. Deviation 5.84
Minimum 1.00
Maximum 34.00
Range 33.00
Interquartile Range 3.95
Skewness 1.67 .37
Kurtosis 5.30 .73

Results with narrative: This category will include above table output as well as following relevant and necessary explanation.

In above table, mean value of hemoglobin for 40 study subjects is 12.24 gm/dl. The ‘Descriptives’ table above also provides 95% Confidence Interval (CI) for the mean estimate. Since 95% CI is basically an interval estimate, its ends are marked by lower and upper bound values. Mean Hb value of 12.24 gm/dl has a 95% CI from 10.37 gm/dl (lower bound) to 14.10 gm/dl (upper bound). Table above also provides values for median (11.55 gm/dl) and 5% trimmed mean (11.76 gm/dl). Latter is a modified mean value of dataset estimated after top 5% and bottom 5% of values are eschewed so as to minimize effect of outliers on mean value. A wide gap between mean (12.24 gm/dl) and trimmed mean (11.76 gm/dl) hints at presence of outlier values in data. This gets further supported by the measurement values of two more useful parameters-minimum and maximum values. The minimum and maximum values of 1 gm/dl and 34 gm/dl clearly suggest occurrence of extreme outliers which may be either a rare phenomenon or typographic error. Also, range and interquartile range are two important measures of capturing dispersion or scatter in a given dataset. Range is simply a value obtained when minimum value is subtracted from maximum value in the same data. Interquartile range (commonly written as IQR) is basically a value obtained as a difference between 75th percentile value (3rd quartile) and 25th percentile value (1st quartile) value. Finally, table (descriptives) contains two more arcane looking statistics called skewness and kurtosis with their coefficient values and corresponding standard errors (abbreviated as ‘std errors’). A value of skewness higher than 0 (preferably (≥ 1) indicates positively skewed data and less than 0 (preferably (≤ 1) indicates negatively skewed data. On the other hand, higher kurtosis coefficient (> +3 in SPSS) indicates positive kurtosis (peakedness of data) and lower kurtosis (<-3) indicates negative kurtosis (flatness of data).