OBS NAME SEGS BAND LYM MON RUN TRIAL 1 BEDDINGFIELD 64 7 20 8 100 0 2 BEDDINGFIELD 66 8 13 13 100 1 3 BEDDINGFIELD 68 6 18 8 400 . . . . 61 4513 69 5 15 7 100 0 62 4513 50 10 32 8 100 1 63 4513 60 8 22 8 400 . Each observation, or row of data, records the results of an analysis of a sample of blood. NAME is the name of the donor. RUN is the number of white blood cells in the sample that were inspected. White blood cells can be classified into four mutually exclusive categories: segmented white blood cells, banded white blood cells, lymphocytes, and monocytes. The variables SEGS, BAND, LYM, and MON denote the percentage of the inspected cells that fall into each category, respectively. Each blood sample is analyzed three separate times as indicated by the variable TRIAL. These data were collected at the pathology laboratory at Rex Hospital. A drop of each donor's blood is placed on a slide. The slide is placed in a centrifuge which spins the slide causing the blood to disperse into a thin layer. Then, the slide is placed in a microscope and the white cells are inspected and classified. These data represent three passes over the same slide, twice at RUN=100 (TRIAL=0,1) and once at RUN=400 (TRIAL=.). This is a new method of slide preparation. The old method was to prepare the blood film by dragging the edge of another slide over the slide with the drop of blood on it. The old method was known to cause an uneven distribution of white cells with, say, proportionately more SEGS in the center than at the edges of the smear. As a consequence, the old method gave different results (different SEGS, BAND, LYM, and MON) for RUN=100 and RUN=400. This was because when RUN=400 a larger area of the slide was scanned to get the larger count and some of the uneven dispersion of white cells got averaged out. These data were collected to determine if this same problem exists with the new method of slide preparation. The assignment is to analyze these data to determine if there is still a difference between results for RUN=100 and RUN=400 with the new method of slide preparation. Present a detailed, careful, complete, and thoughtful analysis. Give attention to the likely distribution function for these data with respect to the assumptions necessary for the validity of your analysis; if corrections are possible to improve the agreement between the data and assumptions, make them. Present your results in the form a report consisting mostly of prose such as you would turn in to a sophisticated client as a statistical consultant. Merely turning in source code and the resulting output is unacceptable.