USER'S GUIDE DUNNETT PROGRAM VERSION 1.5 ECOLOGICAL MONITORING RESEARCH DIVISION ENVIRONMENTAL MONITORING SYSTEMS LABORATORY U. S. ENVIRONMENTAL PROTECTION AGENCY CINCINNATI, OHIO 45268 I. INTRODUCTION This computer program was prepared by under contract with the U.S. Environmental Protection Agency. The program was designed for the analysis of data from acute and short-term chronic toxicity tests with fish and other aquatic life, performed with effluents, receiving waters, and reference toxicants by regulatory agencies and permittees under the National Pollutant Discharge Elimination System, abbreviated NPDES. The toxicity test methods and examples of the statistical analysis of test data are described in detail in the current editions of the following three major EMSL methods manuals: 1. Methods for measuring the acute toxicity of effluents and receiving waters to freshwater and marine organisms. Environmental Monitoring Systems Laboratory, U. S. Environmental Protection Agency, Cincinnati, Ohio 45268. 2. Short-term methods for estimating the chronic toxicity of effluents and receiving waters to freshwater organisms. Environmental Monitoring Systems Laboratory, U. S. Environmental Protection Agency, Cincinnati, Ohio 45268. 3. Short-term methods for estimating the chronic toxicity of effluents and receiving waters to marine and estuarine organisms. Environmental Monitoring Systems Laboratory, U. S. Environmental Protection Agency, Cincinnati, Ohio 45268. Users who are unfamiliar or lack experience with the basic premises of the calculations should carefully read the discussion of Dunnett's Test in the toxicity test manuals listed above and, if necessary, seek the assistance of a statistician before attempting to use this software or interpret the results. It is the user's responsibility to ensure that the data meet the assumptions of the statistical test. Although this user's guide is designed to help the users understand the limitations of the program and how the data must be entered, undoubtedly there will be instances where the results may not appear to be correct. Therefore, the results should be carefully evaluated before use. If you should have any difficulty or questions concerning this software, write or call the Ecological Monitoring Research Division, Environmental Monitoring Systems Laboratory, U.S. Environmental Protection Agency, Cincinnati, Ohio 45268 (Phone: 513-569-7401; Fax: 513-569-7609). II. FILES ON THE DISTRIBUTION DISKETTE: The following files are on the diskette in the subdirectory DUNNETT: A. DUNNETT.BAT - Batch from which the user selects the appropriate program for execution (either DMEAS.EXE or DPROP.EXE.) B. D_MENU.EXE - Executable file which provides DUNNETT.BAT with a menu screen for selecting either DMEAS.EXE or DPROP.EXE. C. DMEAS.EXE - Executable file for measurement data. D. DPROP.EXE - Executable file for proportional data. E. DUNNET.FIL - Contains tables used in Dunnett's Test. F. LABEL - Text file containing program version information. G. README-D.1ST - This user's guide in ASCII text format. H. README-D.WP5 - This user's guide in WordPerfect 5.1 format. I. TESTP-D.INP - Example data file for batch input of proportional data. J. TESTM-D.INP - Example data file for batch input of measurement data. III. INSTALLATION A. Make a backup copy of the DUNNETT program files on the distribution diskette, and store the original diskette in a safe place. B. The DUNNETT program can be executed directly from a floppy disk, if necessary, or from the hard disk. To execute from the hard disk, create all appropriate (sub) directories on the hard disk and use the DOS "COPY" command to copy the files from the diskette to the new (sub) directories. IV. PROGRAM OVERVIEW The DUNNETT program components are written in Microsoft FORTRAN Version 5.1 for DOS on IBM-compatible PCs. They provide for real-time keyboard data input, or batch input via previously prepared data files. The multiple comparison is based on Dunnett, C. W., 1955, "Multiple Comparison Procedure for Comparing Several Treatments with a Control," J. Amer. Statist. Assoc. 50:1096-1121. The software performs the following functions: A. Performs an analysis of variance (ANOVA), which is used to obtain the error value for Dunnett's Procedure. B. Performs a multiple comparison of treatment means with the control mean (Dunnett's Procedure). Dunnett's Procedure indicates which toxicant concentration means (if any) are statistically different from the control mean at the 5% level of significance. C. Calculates the minimum difference between the control and treatment means that could be detected as statistically significant, and tests the validity of the homogeneity of variance assumption using Bartlett's Test. V. PROGRAM RESTRICTIONS The following restrictions apply to the DUNNETT program: A. No less than 2 and no more than 8 treatments (concentrations), including the control, can be used. B. No more than 50 replicates can be used per concentration. C. The number of degrees of freedom for error must be 5 or greater. D. Input must come from a data file. Data files can be created either within the program or externally using a text editor. VI. PROGRAM EXECUTION You have been supplied with a compiled version of the program. To execute the program, from the appropriate (sub) directory DOS prompt (i.e. A:\DUNNETT or C:\DUNNETT) type: DUNNETT and press the "enter" key. A. Program Input and Execution Options 1. When the program is initiated, you are asked first to designate whether the data to be entered or analyzed is proportional data (such as percent survival) or measurement data (such as weights, numbers of offspring, etc.) The program has separate data entry and analysis routines for the two types of data. Once you select one type of data, YOU MUST EXIT THE PROGRAM AND RESTART IT TO CHANGE DATA TYPES. There is no way to change from one type of data to the other within the program itself. 2. You are then asked to input a "Title" and to choose between output to a printer or a disk file. If output is directed to a disk file, you are prompted to provide a complete filename, including extension. 3. The program then presents the user with the following options: a. Create a data file b. Edit a data file c. Analyze an existing data set d. Stop (exit the program) Examples of the uses of the options are provided in the examples section below. 4. If the option "Create a data file" is selected and the data are input, they must be saved to a file before using in a statistical analysis. If you do not specify a "path name" for the data file when saving it, the file will be saved to the main directory in which the DUNNETT program files are stored, the default directory. If you want the data file to be stored in a different directory, you must specify the "path name" as well as the file name. IMPORTANT: Data input to the screen can not be used directly for Dunnett's Procedure. The data must be saved to a file, and called up for analysis by selecting the option "Analyze an existing data set." Real-time data input to the screen can be tedious and time consuming for some samples. The user may find it more convenient to prepare data files outside the DUNNETT program using an ASCII test editor. Sample formats for data files can be found in section VII-C of this document. The new data file can then be retrieved and checked with the "Edit a data file" option before it is analyzed. 5. If the option "Edit a data file" is selected, the program asks for the file name, and then the (ID) number of the toxicant concentration (1-8) for which data are to be edited. In editing proportional data, the program asks for a change in a specific replicate number. In editing measurement data, however, the program does not ask for changes by replicate. The user is prompted to change the values within a specific concentration, without regard to the order of the individual replicates. If there are multiple occurrences of a specific value, any change will be made to the first occurrence of that value. Again, note that the program does not require the values to be in any certain order within a concentration group. When the editing process is complete, the data must be saved to a file before it can be analyzed. The edited data are saved under the file name used for retrieval. Examples of data editing can be found in section VII-D of this document. NOTE: The user does NOT have an option to change the name of the edited file before saving it. 6. If the option "Analyze an existing data set" is selected, the program asks for the data file name. 7. If the option "Stop" is selected, the program is terminated. Note that you may halt execution of the program at any time by pressing the "CTRL" key and the "BREAK" key at the same time. B. Data Transformations When the option "Analyze a data set" is selected, the program requests the data file name and presents the user with a set of options for data transformation: 1. No transformation 2. Square root 3. Log 10 4. Arc sine square root NOTE: The arc sine square root transformation is available only in the program sections dealing with proportional data. Consult the toxicity test methods manuals for the correct selection. The transformed data are not saved. C. Data Analysis Following transformation (if any) of the data, the following analyses are completed: 1. A one-way analysis of variance. 2. Bartlett's test for homogeneity of variance. 3. If the numbers of replicates are the same for all concentrations, Dunnett's Test is used to compare the control mean with each of the concentration means. 4. If the number of replicates is not the same for all concentrations, simple t-tests, using a pooled error term, are performed with Bonferroni's adjustment of the alpha level. 5. The minimum significant difference that can be detected from the control group mean is calculated. D. Program Output The program output includes the following: 1. Separate tables of summary statistics for the raw data and transformed data (if a transformation is selected.) The data include the mean, standard deviation and coefficient of variation of the responses at each concentration level. Press the "return" key to advance the screen from one table to the next during program execution. 2. The results of the multiple comparison procedure. Concentration means which are significantly different from the control are indicated on the summary table (alpha=0.05, 1-sided) with asterisks. 3. The minimum significant (detectable) difference from the control group mean. 4. Miscellaneous statistics: a. Between-concentration sum of squares b. Error mean square c. Bartlett's test p-value for equality of variances 5. Warning Messages The program generates warning messages related to data input and analysis. Messages related to data analysis are: a. "The test for equality of variances is significant (p less than 0.01). The results of this analysis should be interpreted with caution." b. "The test for equality of variances could not be computed because one or more of the variances is zero." These warning messages indicate that the data do not satisfy the homogeneity of variance assumption of Dunnett's Procedure. In these cases, use of a nonparametric procedure such as Steel's Many-one Rank Test is recommended. VII. EXAMPLES OF DATA ANALYSIS A. Data on the percent survival of Mysidopsis bahia from a short-term chronic survival, growth, and fecundity test are used to illustrate the data input and analysis output for the proportional data sections of the program. The survival data, expressed as numbers of survivors, are summarized in Table 1. Examples of program input and output are provided in Figures 1 and 2, respectively. TABLE 1. SAMPLE DATA FOR PROPORTION SECTION OF DUNNETT PROGRAM: NUMBER OF MYSIDS SURVIVING (5 ORGANISMS PER REPLICATE) _________________________________________________________________________ NUMBER OF SURVIVORS ________________________________________________________ REPLICATE TOXICANT ________________________________________________________ (EFFLUENT) CONCENTRATION 1 2 3 4 5 6 7 8 _________________________________________________________________________ 1 (CONTROL) 4 4 4 5 5 5 5 5 2 (6.25%) 5 5 5 5 5 3 3 3 3 (12.5%) 5 3 3 5 3 5 3 3 4 (25.0%) 3 0 4 5 5 5 3 3 5 (50.0%) 1 3 5 5 0 1 3 0 6 (100.0%) 0 3 1 3 0 3 3 3 _________________________________________________________________________ FIGURE l. EXAMPLE OF DUNNETT PROGRAM PROPORTIONAL DATA INPUT FROM THE SCREEN AND ANALYSIS USING THE DATA IN TABLE 1 ______________________________________________________________________________ EMSL Cincinnati Dunnett Software Version 1.5 What type of data do you wish to analyze? 1) response proportions (like survival data or fertility proportion data) Note: The program calculates a proportion after prompting for number of exposed organisms and number of responding organisms. 2) counts and measurements (like offspring counts, cystocarps and algal cell counts, weights, chlorophyll measurements, or turbidity measurements) Enter "1", "2", (or "q" to quit program): 1 Title ? Mysid Survival Data Output to printer or disk file ? P 1) Create a data file 2) Edit a data file 3) Analyze an existing data set 4) Stop Your choice ? 1 ______________________________________________________________________________ FIGURE l. EXAMPLE OF DUNNETT PROGRAM PROPORTIONAL DATA INPUT FROM THE SCREEN AND ANALYSIS USING THE DATA IN TABLE 1 (CONTINUED) ______________________________________________________________________________ Number of concentrations, including control ? 6 Number of replicates for conc. 1 (the control) ? 8 replicate number of organisms exposed number of organisms responding (organisms surviving, eggs fertilized, etc.) 1 5 4 2 5 4 3 5 4 4 5 5 5 5 5 6 5 5 7 5 5 8 5 5 Number of replicates for conc. 2 ? 8 replicate number of organisms exposed number of organisms responding (organisms surviving, eggs fertilized, etc.) 1 5 5 2 5 5 3 5 5 4 5 5 5 5 5 6 5 3 7 5 3 8 5 3 Number of replicates for conc. 3 ? 8 replicate number of organisms exposed number of organisms responding (organisms surviving, eggs fertilized, etc.) 1 5 5 2 5 3 3 5 3 4 5 5 5 5 3 6 5 5 7 5 3 8 5 3 ______________________________________________________________________________ FIGURE l. EXAMPLE OF DUNNETT PROGRAM PROPORTIONAL DATA INPUT FROM THE SCREEN AND ANALYSIS USING THE DATA IN TABLE 1 (CONTINUED) ______________________________________________________________________________ Number of replicates for conc. 4 ? 8 replicate number of organisms exposed number of organisms responding (organisms surviving, eggs fertilized, etc.) 1 5 3 2 5 0 3 5 4 4 5 5 5 5 5 6 5 5 7 5 3 8 5 3 Number of replicates for conc. 5 ? 8 replicate number of organisms exposed number of organisms responding (organisms surviving, eggs fertilized, etc.) 1 5 1 2 5 3 3 5 5 4 5 5 5 5 0 6 5 1 7 5 3 8 5 0 Number of replicates for conc. 6 ? 8 replicate number of organisms exposed number of organisms responding (organisms surviving, eggs fertilized, etc.) 1 5 0 2 5 3 3 5 1 4 5 3 5 5 0 6 5 3 7 5 3 8 5 3 ______________________________________________________________________________ FIGURE l. EXAMPLE OF DUNNETT PROGRAM PROPORTIONAL DATA INPUT FROM THE SCREEN AND ANALYSIS USING THE DATA IN TABLE 1 (CONTINUED) ______________________________________________________________________________ Do you wish to save the data on disk ?Y Disk file for output ? (Enter complete DOS file name.) c:\dunnett\mysid.dat 1) Create a data file 2) Edit a data file 3) Analyze an existing data set 4) Stop Your choice ? 3 Data file to be analyzed ? (Enter complete DOS file name.) c:\dunnett\mysid.dat Available Transformations 1) no transform 2) square root 3) log10 4) arc sine square root Your choice ? 4 Dunnett's test as implemented in this program is a one-sided test. You must specify the direction the test is to be run; that is, do you expect the means for the test concentrations to be less than or greater than the mean for the control concentration. Direction for Dunnetts test : L=less than, G=greater than ? L ______________________________________________________________________________ FIGURE 2. EXAMPLE OF DUNNETT PROGRAM PROPORTIONAL DATA ANALYSIS OUTPUT USING THE DATA INPUT IN FIGURE 1 ______________________________________________________________________________ Mysid Survival Data Summary Statistics for Raw Data Conc. n Mean s.d. cv% ____________________________________________________________ 1 = control 8 .9250 .1035 11.2 2 8 .8500 .2070 24.4 3 8 .7500 .2070 27.6 4 8 .7000 .3381 48.3 5 8 .4500 .4106 91.2 6 8 .4000 .2828 70.7 ____________________________________________________________ Mysid Survival Data Summary Statistics and ANOVA Transformation = Arc Sine Square Root Conc. n Mean s.d. cv% ____________________________________________________________ 1 = control 8 1.2560 .1232 9.8 2 8 1.1731 .2377 20.3 3 8 1.0583 .2377 22.5 4 8 1.0033 .3794 37.8 5* 8 .7301 .4562 62.5 6* 8 .6681 .3096 46.3 ____________________________________________________________ *) the mean for this conc. is significantly less than the control mean at alpha = 0.05 (1-sided) by Dunnett's test Minimum detectable difference for Dunnett's test = -.357901 This corresponds to a difference of -.292391 in original units This difference corresponds to -32.34 percent of control Between concentrations sum of squares = 2.238431 with 5 degrees of freedom. Error mean square = .096020 with 42 degrees of freedom. Bartlett's test p-value for equality of variances = .041 ______________________________________________________________________________ B. Reproduction of Ceriodaphnia dubia from a short-term chronic survival and reproduction test are used to illustrate the program data input and analysis output for the measurement data sections of the program. The reproduction data, expressed as numbers of offspring, are summarized in Table 2. Examples of program input and output are provided in Figures 3 and 4, respectively. TABLE 2. SAMPLE DATA FOR MEASUREMENT DATA SECTION OF DUNNETT PROGRAM: CERIODAPHNIA DUBIA REPRODUCTION DATA ____________________________________________________________________________ NO. YOUNG PER ADULT __________________________________________________ REPLICATE TOXICANT __________________________________________________ NO. (EFFLUENT) LIVE CONCENTRATION 1 2 3 4 5 6 7 8 9 10 ADULTS ____________________________________________________________________________ 1 (CONTROL) 27 30 29 31 16 15 18 17 14 27 10 2 (1.56%) 32 35 32 26 18 29 27 16 35 13 10 3 (3.12%) 39 30 33 33 36 33 33 27 38 44 10 4 (6.25%) 27 34 36 34 31 27 33 31 33 31 10 5 (12.5%) 10 13 7 7 7 10 10 16 12 2 10 6 (25.0%)* 0 0 0 0 0 0 0 0 0 0 3 ____________________________________________________________________________ *Since there is significant mortality in the 25% effluent concentration, its effect on reproduction is not considered. FIGURE 3. EXAMPLE OF DUNNETT PROGRAM MEASUREMENT DATA INPUT FROM THE SCREEN AND ANALYSIS USING THE DATA IN TABLE 2 ______________________________________________________________________________ EMSL Cincinnati Dunnett Software Version 1.5 What type of data do you wish to analyze? 1) response proportions (like survival data or fertility proportion data) Note: The program calculates a proportion after prompting for number of exposed organisms and number of responding organisms. 2) counts and measurements (like offspring counts, cystocarps and algal cell counts, weights, chlorophyll measurements, or turbidity measurements) Enter "1", "2", (or "q" to quit program): 2 Title ? Ceriodaphnia dubia Reproduction Data Output to printer or disk file ? P 1) Create a data file 2) Edit a data file 3) Analyze an existing data set 4) Stop Your choice ? 1 Number of concentrations, including control ? 5 ______________________________________________________________________________ FIGURE 3. EXAMPLE OF DUNNETT PROGRAM MEASUREMENT DATA INPUT FROM THE SCREEN AND ANALYSIS USING THE DATA IN TABLE 2 (CONTINUED) ______________________________________________________________________________ Number of observations for conc. 1 (the control) ? 10 Enter the data for conc. 1 (the control) one observation at a time. N0. 1? 27 N0. 2? 30 N0. 3? 29 N0. 4? 31 N0. 5? 16 N0. 6? 15 N0. 7? 18 N0. 8? 17 N0. 9? 14 N0. 10? 27 Number of observations for conc. 2 ? 10 Enter the data for conc. 2 one observation at a time. N0. 1? 32 N0. 2? 35 N0. 3? 32 N0. 4? 26 N0. 5? 18 N0. 6? 29 N0. 7? 27 N0. 8? 16 N0. 9? 35 N0. 10? 13 ______________________________________________________________________________ FIGURE 3. EXAMPLE OF DUNNETT PROGRAM MEASUREMENT DATA INPUT FROM THE SCREEN AND ANALYSIS USING THE DATA IN TABLE 2 (CONTINUED) ______________________________________________________________________________ Number of observations for conc. 3 ? 10 Enter the data for conc. 3 one observation at a time. N0. 1? 39 N0. 2? 30 N0. 3? 33 N0. 4? 33 N0. 5? 36 N0. 6? 33 N0. 7? 33 N0. 8? 27 N0. 9? 38 N0. 10? 44 Number of observations for conc. 4 ? 10 Enter the data for conc. 4 one observation at a time. N0. 1? 27 N0. 2? 34 N0. 3? 36 N0. 4? 34 N0. 5? 31 N0. 6? 27 N0. 7? 33 N0. 8? 31 N0. 9? 33 N0. 10? 31 ______________________________________________________________________________ FIGURE 3. EXAMPLE OF DUNNETT PROGRAM MEASUREMENT DATA INPUT FROM THE SCREEN AND ANALYSIS USING THE DATA IN TABLE 2 (CONTINUED) ______________________________________________________________________________ Number of observations for conc. 5 ? 10 Enter the data for conc. 5 one observation at a time. N0. 1? 10 N0. 2? 13 N0. 3? 7 N0. 4? 7 N0. 5? 7 N0. 6? 10 N0. 7? 10 N0. 8? 16 N0. 9? 12 N0. 10? 2 Do you wish to save the data on disk ?Y Disk file for output ? c:\dunnett\cerio.dat 1) Create a data file 2) Edit a data file 3) Analyze an existing data set 4) Stop Your choice ? 3 File name ? c:\dunnett\cerio.dat ______________________________________________________________________________ FIGURE 3. EXAMPLE OF DUNNETT PROGRAM MEASUREMENT DATA INPUT FROM THE SCREEN AND ANALYSIS USING THE DATA IN TABLE 2 (CONTINUED) ______________________________________________________________________________ Available Transformations 1) no transform 2) square root 3) log10 Your choice ? 1 Dunnett's test as implemented in this program is a one-sided test. You must specify the direction the test is to be run; that is, do you expect the means for the test concentrations to be less than or greater than the mean for the control concentration. Direction for Dunnetts test : L=less than, G=greater than ? L ______________________________________________________________________________ FIGURE 4. EXAMPLE OF DUNNETT PROGRAM MEASUREMENT DATA ANALYSIS OUTPUT USING THE DATA INPUT IN FIGURE 1 ______________________________________________________________________________ Ceriodaphnia dubia Reproduction Data Summary Statistics and ANOVA Transformation = None Conc. n Mean s.d. cv% ____________________________________________________________ 1 = control 10 22.4000 6.9314 30.9 2 10 26.3000 8.0007 30.4 3 10 34.6000 4.8351 14.0 4 10 31.7000 2.9458 9.3 5* 10 9.4000 3.8930 41.4 ____________________________________________________________ *) the mean for this conc. is significantly less than the control mean at alpha = 0.05 (1-sided) by Dunnett's test Minimum detectable difference for Dunnett's test = -5.628560 This difference corresponds to -25.13 percent of control Between concentrations sum of squares = 3887.880000 with 4 degrees of freedom. Error mean square = 31.853333 with 45 degrees of freedom. Bartlett's test p-value for equality of variances = .029 Do you wish to restart the program ? N ______________________________________________________________________________ C. Formats for creation and/or editing of ASCII files: The examples below provide the file formats for ASCII files which may be used instead of data input via the screen. The files have a columnar format in which the column delimiters may be either commas or blanks. Note that the columns do not have to be a specific width. In these files it is assumed that the group identified as group 1 contains data for the control group. Note that the file groups may have unequal sample sizes. Examples of data input from externally created ASCII files are provided in Figures 5 and 6. 1. Proportion data These files consist of three columns. The first column identifies the group of replicates. The second column identifies number exposed for each replicate and the third column indicates number responding for each replicate. The statistical calculations are performed on proportions consisting of the number responding divided by number exposed. The data below represents a portion of the file TESTP-D.INP which contains the mysid survival data from Example A, above. It may be used as a template to create data files for analysis with the DUNNETT program, to avoid data input via the screen. TABLE 3. SAMPLE FORMAT FOR PROPORTIONAL DATA INPUT ASCII FILES _______________________________________________________________ 1, 5, 4 1, 5, 4 1, 5, 4 1, 5, 5 1, 5, 5 1, 5, 5 1, 5, 5 1, 5, 5 2, 5, 5 2, 5, 5 2, 5, 5 2, 5, 5 2, 5, 5 2, 5, 3 2, 5, 3 2, 5, 3 _______________________________________________________________ 2. Measurement data and counts These files consist of two columns. The first column denotes the group of replicates and the second column the data value. The data below represents a portion of the file TESTP-D.INP which contains the Ceriodaphnia dubia reproduction data from Example B, above. It may be used as a template to create data files for analysis with the DUNNETT program, to avoid data input via the screen. TABLE 4. SAMPLE FORMAT FOR MEASUREMENT DATA INPUT ASCII FILES _______________________________________________________________ 1, 27 1, 30 1, 29 1, 31 1, 16 1, 15 1, 18 1, 17 1, 14 1, 27 2, 32 2, 35 2, 32 2, 26 2, 18 2, 29 2, 27 2, 16 2, 35 2, 13 _______________________________________________________________ FIGURE 5. EXAMPLE OF DUNNETT PROGRAM PROPORTIONAL DATA INPUT FROM AN ASCII FILE WHICH CONTAINS THE DATA IN TABLE 1 (PORTIONS OF SCREEN PROMPTS AND RESPONSES HAVE BEEN OMITTED) ______________________________________________________________________________ Title ? Mysid Survival Data from file TESTP-D.INP Output to printer or disk file ? D Disk file for output ? TESTP-D.OUT 1) Create a data file 2) Edit a data file 3) Analyze an existing data set 4) Stop Your choice ? 3 File name ? C:\DUNNETT\TESTP-D.INP ______________________________________________________________________________ FIGURE 6. EXAMPLE OF DUNNETT PROGRAM MEASUREMENT DATA INPUT FROM AN ASCII FILE WHICH CONTAINS THE DATA IN TABLE 2 (PORTIONS OF SCREEN PROMPTS AND RESPONSES HAVE BEEN OMITTED) ______________________________________________________________________________ Title ? Ceriodaphnia dubia Reproduction Data from file TESTM-D.INP Output to printer or disk file ? D Disk file for output ? TESTM-D.OUT 1) Create a data file 2) Edit a data file 3) Analyze an existing data set 4) Stop Your choice ? 3 File name ? C:\DUNNETT\TESTM-D.INP D. Program data editing facilities The examples below illustrate the DUNNETT program's data editing facilities, which differ for proportion and measurement data. In Figure 7, one observation from an ASCII file containing the proportion survival data from Table 1 is altered and then changed back to its original value. Figure 8 shows a similar alteration of an observation from an ASCII file containing the reproduction data from Table 2. FIGURE 7. EXAMPLE OF DUNNETT PROGRAM PROPORTIONAL DATA FILE EDIT ROUTINE OF AN ASCII FILE WHICH CONTAINS THE DATA IN TABLE 1. ONE OBSERVATION IS ALTERED AND THEN CHANGED BACK. (PORTIONS OF SCREEN PROMPTS AND RESPONSES HAVE BEEN OMITTED) ______________________________________________________________________________ 1) Create a data file 2) Edit a data file 3) Analyze an existing data set 4) Stop Your choice ? 2 data file to be edited ? (Enter complete DOS file name.) C:\DUNNETT\TESTP-D.INP Concentration being edited ? (Enter a concentration number or "e" to end editing.) 1 replicate number of organisms exposed number of organisms responding (organisms surviving, eggs fertilized, etc.) 1 5 4 2 5 4 3 5 4 4 5 5 5 5 5 6 5 5 7 5 5 8 5 5 replicate being edited ? (Enter a replicate number or "e" to end editing this conc.) 5 number of organisms exposed ? 5 number of organisms responding ? 3 ______________________________________________________________________________ FIGURE 7. EXAMPLE OF DUNNETT PROGRAM PROPORTIONAL DATA FILE EDIT ROUTINE OF AN ASCII FILE WHICH CONTAINS THE DATA IN TABLE 1. ONE OBSERVATION IS ALTERED AND THEN CHANGED BACK. (PORTIONS OF SCREEN PROMPTS AND RESPONSES HAVE BEEN OMITTED) (CONTINUED) ______________________________________________________________________________ replicate number of organisms exposed number of organisms responding (organisms surviving, eggs fertilized, etc.) 1 5 4 2 5 4 3 5 4 4 5 5 5 5 3 6 5 5 7 5 5 8 5 5 replicate being edited ? (Enter a replicate number or "e" to end editing this conc.) 5 number of organisms exposed ? 5 number of organisms responding ? 5 replicate number of organisms exposed number of organisms responding (organisms surviving, eggs fertilized, etc.) 1 5 4 2 5 4 3 5 4 4 5 5 5 5 5 6 5 5 7 5 5 8 5 5 replicate being edited ? (Enter a replicate number or "e" to end editing this conc.) e Do you wish to save the changes ? (Y/N) Y ______________________________________________________________________________ FIGURE 8. EXAMPLE OF DUNNETT PROGRAM MEASUREMENT DATA FILE EDIT ROUTINE OF AN ASCII FILE WHICH CONTAINS THE DATA IN TABLE 2. ONE OBSERVATION IS ALTERED AND THEN CHANGED BACK. (PORTIONS OF SCREEN PROMPTS AND RESPONSES HAVE BEEN OMITTED) ______________________________________________________________________________ 1) Create a data file 2) Edit a data file 3) Analyze an existing data set 4) Stop Your choice ? 2 File name ? C:\DUNNETT\TESTM-D.INP Edit values for which concentration ? 2 The following values are for conc. 2 32.000000 35.000000 32.000000 26.000000 18.000000 29.000000 27.000000 16.000000 35.000000 13.000000 Modify a value, Edit a different conc. or Quit (M,E,Q) ? M Enter old value, new value: 29, 31 32.000000 35.000000 32.000000 26.000000 18.000000 31.000000 27.000000 16.000000 35.000000 13.000000 Modify a value, Edit a different conc. or Quit (M,E,Q) ? M Enter old value, new value: 31, 29 The following values are for conc. 2 32.000000 35.000000 32.000000 26.000000 18.000000 29.000000 27.000000 16.000000 35.000000 13.000000 Modify a value, Edit a different conc. or Quit (M,E,Q) ? Q Do you wish to save the changes ? Y ______________________________________________________________________________