You've already forked lazarus-ccr
LazStats: Delete the file source/LazStats.ini.
git-svn-id: https://svn.code.sf.net/p/lazarus-ccr/svn@7827 8e941d3f-bd1b-0410-a28a-d453659cc2b4
This commit is contained in:
wp_xxyyzz
@ -1,64 +0,0 @@
|
||||
[LANGUAGE]
|
||||
DEFAULT=ENGLISH
|
||||
[ENGLISH]
|
||||
101=Directions: For entry of data on this form, enter the number of rows, columns and slices in the boxes provided. Press enter after each entry. Then enter the frequencies observed for each cell in the grid. If entering data from a file in the main grid, select the row, column, slice and frequency variables by selecting from the list of variables and clicking the right-arrow for the corresponding variable. Click the compute button to obtain the results.
|
||||
102=The AxBxR ANOVA involves two between treatment factors and repeated measures factors. Two grid column variables contain the A and B treatment values (codes 1, 2, etc.) and 2 to K grid column variables for the repeated measure observations. All ABC groups are assumed to be of the same size. There is a maximum of 20 repeated measures.
|
||||
103=This procedure analyzes fixed effects with up to three levels of interaction and one or more covariates. Multiple regression methods are used (See "Multiple Regression in Behavioral Research" by Elazar J. Pedhazur, Harcourt, Brace, College Publishers, 1997, Chapter 16, pages 675-713.) A test is performed for the assumption of homogeneous regression slopes in addition to the ANCOVA. Both adjusted and unadjusted means are reported. Comparisons are made among the adjusted means.
|
||||
104=Directions: Select a variable to analyze. You may analyxe series from either a column (default) or a "Case" row. You may elect to analyze all values in a column (or row) as desiredClick the buttons for any desired smoothing options. The program will automatically "split" the list of row values (or column values) for that variable into two sub-sets of X and Y scores with each Y score being the value which "lags" behind the X score in the list by k lag values. All possible lags which yield a sample as large as 3 or more are computed and plotted in a "correlogram". You may optionally print the lag,
|
||||
correlation, means, standard deviations and confidence interval for each correlation. The differences between original and smoothed values (residuals) may be plotted. The smoothed points replace the original values in the analysis if smoothing is elected.
|
||||
105=The main grid should contain a symetric matrix of similarity or dissimilarity values representing distances among the objects to be clustered. Check the type box io indicate if the measures are similarities (e.g. correlations) or dissimilarities.
|
||||
106=Directions: It is assumed you have one grid column variable representing the group codes for the (A) between treatment groups effect and 2 to k column variables representing the repeated measures. Group codes should be sequential values of 1, 2, etc. You may elect to plot the means.
|
||||
107=Directions: You may elect to complete a 1, 2 or 3 way ANOVA by selecting a dependent variable and 1, 2 or 3 factor variables. If you elect post-hoc tests, comparisons are made between factor levels. NOTE: some post-hoc comparisons are made only with equal N's.
|
||||
108=Directions: First click on the variable name that represents the group codes. Next, click on the variable that reresents the measurement to be plotted. Click the Compute button to obtain the results. You can obtain a single boxplot for all cases if you use a "dummy" group variable containing only the group code 1 for all cases.
|
||||
109=Directions: First, select the categorical variables of your study. Select them in the order of the desired breakdown. Next, slect the continuous variable for which you want statistics for each cell obtained by by breaking down the categorical variables into their respective categories.
|
||||
110=Directions: 1. Select the variable containing the bubble identification mumber - an integer in the range of 1 to N objects. 2. Select the variable representing the X axis integer value for the object. This is the repeated measures variable. 3. Select the variable representing the Y axis. This should be a floating point value. 4. Select the variable representing the size of the bubble for each object to be plotted at the X and Y location. NOTE: Each data line reresents one replication (X value) of the object plotted. See the example data in the file labeled BubblePlot.LAZ
|
||||
111=NOTE: No. of left hand variables must be less than or equal to the number of right hand variables.
|
||||
112=Directions: Click on the variable that represents the measurement. Click on the Sigma button to change the default value. Click the Compute button to obtain the results.
|
||||
113=Directions: Two to k variables representing dichotomous (0,1) values are analyzed for N cases. The values of the variables reflect repeated observations on the same subjects or on matched subjects. Click the variables on the left to analyze and enter them by clicking the right arrow button.
|
||||
114=Directions: Forst select the test scores from the available variables. You will see a default reliability and weight ssigned to each score selected in list boxes to the right. If you click on either a reliability or a weight, an input box will appear in which you can enter a new reliability or weight. Note - you can use the KR#21 reliability program to estimate reliability if you know the maximum score.
|
||||
115=Directions: A Judge's ratings or observations are recorded as variables (columns) 1 through k. Each line conrresponds to a different judge (person making the rating.) Select the variables from the left list to analyze and click on the right arrow button to enter them. To remove a variable from the list of selected ones, click on the variable name in the selected list and click the left arrow button. Click the Compute button to obtain the results.
|
||||
116=Directions: Select each categorical variable from the available variables in the leftmost box in the order that you wish to have the breakdown proceed. Click the OK button to start the analysis.
|
||||
117=Directions: First, click on the variable name that represents the sample lot number. Next, click on the variable that represents the measurement. Click on the Delta size and enter the desired value. Click on the alpha and/or beta probability boxes and enter values to change from the default values. You may also enter target specifications if you first click the check box to use a target specification.
|
||||
118=Description: Double Declining Value determines accelerated depreciation values for an asset given the initial cost, life expectancy, and value, and depreciation period. EXAMPLE: What is the depreciation value for a computer with a life expectancy of three years if it initially cost $2,000.00 with no expected value at the end of the three years? Initial Cost = 2000.00 Life Expectancy = 3 End Value = 0.0 Depreciation Period = 3 ANSWER: $148.15
|
||||
119=This procedure provides means, variances, standard deviations, skewness, kurtosis and range values for each variable selected. Select the variables in the left list and enter them for analysiis by clicking the right arrow. If you select the z score option, a new variable will be added to your grid for each variable you select. The new variable will contain the transformation of the original variable into a z score.
|
||||
120=Each row of the grid below corresponds to one column of the data grid. Complete the information requested in each cell of the row. To add another variable (row in the dictionary), press the down-arrow on your keyboard.
|
||||
121=Directions: Specify the lag value for the differences desired, e.g. 1 to obtain the difference between point 1 and 2, 2, and 3, etc. Also, indicate the order, i.e. the number of times to repeat the differencing operation. Click OK when ready.
|
||||
122=This procedure is an adaptation of the program written by Niels G. Waller, Dept. of Psychology, University of California-Davis, Jan. 1998. It's purpose is to identify test items that differ in the response pattern for two groups: a reference group and a focal group. The file of data to be analyzed should consist of a variable containing a code designating the two groups and variables containing subject's item responses coded 0 for incorrect and 1 for correct. No missing data may be included. The results provide the Mantel-Haenszel statistics for identifying those items which are different for the two groups.
|
||||
123=Directions: The number of intervals may not exceed the number of cses. To change the interval size, click on the current size and replace it with a new size. Press the enter button after entering a new value.
|
||||
124=Directions: The two way ANOVA on ranks is similar to a mixed design ANOVA with repeated measures (1 to k conditions) on ssubjects in 1 to M groups. The program expects one variable to represent the group code, and 1 to k score variables for each case. The scores for the cases in each group are used to obtain rankings among the k scores within each group. The test is whether or not the rank totals for the conditions are equal within the expected sampling variability.
|
||||
125=This procedure calculates the Kappa coefficient for objects or subjects classified into two or more categories by a group of judges or procedures. Each object is coded with a sequential integer ranging from 1 to the number of objects. Each judge is also coded with an integer from 1 to the number of judges. Categories are numbered with integers from 1 to the number of categories. These are column variables. It is expected that the total number of cases will be the number of judges times the number of objects.
|
||||
126=Directions: he GLM procedure permits the user to specify multiple dependent variables and multiple independent variables. Variables for both dependent and independent may be either continuous or categorical variables. The independent variables are classified as fixed effects, random effects, repeated measures or covariates. Interactions among the independent variables may be specified for the model used. To define an interaction in your model, click the start definition button and then click on each independent variable to be included in the interaction. Click the end definition button to end the definition. A maximum of 5 terms is allowed in an interaction.
|
||||
127=You may obtasin results for a single group or for experimental and control groups. If there is only one group, leave the group variable blank. Data entered on each line of the data grid represent one case within a group. You will typically have two or three columns of data with variable labels like "TIME", "GROUP" and "EVENT". Each variable should be defined as an integer in your variable definitions. Note that the code for experimental and control groups are 1 and 2. The coding for the event or censored is 1 for the event (death) and 2 for the censored (lost, can't observe.) An example file with the name "KaplanMeierTest.LAZ is available for use.
|
||||
128=This procedure provides both the weighted and unweighted Kappa Coefficients for assessing the consistency of judgements for two raters. It also provides other measures of the independence of the ratings. If nominal categories are used in the ratings, the unweighted statistic is appropriate. If the categories represent ordinal data, the weighted Kappa statistics may be appropriate. The number of rows must equal the number of columns to calculate the Kappa statistics.
|
||||
129=The main grid should contain data values representing variables meansured on the objects to be clustered (rows.) Enter the desired number of clusters, select the variables to use in clustering and select the options desired.
|
||||
130=See B. J. Winer's "Statistical Principles in Experimental Design", McGraw-Hill Book Company, New York, 1962, pages 514-577 for the analyses plans provided in this procedure. Note: Factor codes should be formatted as integers, data values as floating point values. All cell sizes should be equal and no missing values are allowed.
|
||||
131=Complete the specifications for your log-linear analysis of cross-classivation data as indicated below. Complete step 1, step 2 and step 3. Select any options desired. Click the Compute button to obtain the results. Should you need to start over, click the Reset button. When your analysis or analyses are commpleted, click the Return button.
|
||||
132=Directions: Enter the order of the moving averae. The order is the number of values on each side of a point to be included in the average. When you enter a value, a list of corresponding thetas will appear in the list. Click on each theta of the list for entry of the desired weight (default 1.0). Enter a weight in the theta value box and press the enter key. Repeat for each theta in the list. Click the Apply button when ready. The theta values will be re-proportioned to sum to 1.0 accross all values. Click the OK button to continue.
|
||||
133=Directions: You may generate sample multivariate data from a population with known intercorrelation among the variables and with known population means and standard deviations. Enter the number of variables and size of the sample to generate. Then enter the correlations among the variables row-wise the program will fill in the lower triangular values.) Next, enter the population means and standard deviations. When ready to generate the data, click the ComputeBtn. The data will be placed in the data grid. You can save this data to a file.
|
||||
134=Directions: 1. Select the X Variable 2. Select the Y variable 3. Select the group variable (integer) 4. Enter a label for the plot 5. Select an option if desired. 6. Click the Compute button
|
||||
135=If you use a language that uses the comma (,) separator to separate the whole part from the fractional part of a number (e.g. 123,45) then select the EUROPEAN option. The default is the English convention (a period, e.g. 123.45). You can enter a default directory to locate your data files. Click a button for values that represent a missing value and click a button that indicates how you want to display values in a grid cell (justification.)
|
||||
136=For partial and semi-partial correlations, select the dependent variable then select the predictor variable(s), and finally the variable(s) to be partialled. Note that simple, higher order and multiple simple and higher order partialling may be completed as a function of the number of predictors and partialled variables included in the analysis.
|
||||
137=Directions: The p Chart for nonconforming parts assumes you have a variable (column of data) which represents the number of nonconforming parts in a sample lot of size N. You are expected to enter the sample size N in which each of the observations was made. You will also need to enter P, the expected or target proportion of defects in a sample of N parts. To select the measurement variable, click on the name of the variable in the list of variables available. Enter the N and P values in the boxes provided. If you desire a sigma value other than the default, click the desired button. Click the Compute button to obtain the results.
|
||||
138=Directions: To use the program you should have the following values coded for each subject: (1) a variable (1/2) for the reference or focus group. (2) one or more items which contain an item score (integers representing response categories, e.g. a value from 1 to 5. Follow these steps to complete the analysis: (1) Enter the items in the available variables list into the selected items list. (2) Enter the group variable from the available variables list to the group box. (3) Enter the Lowest Item Score in the corresponding box. (4) Enter the highest Item Score in the corresponding box. (5) Enter the Reference Group Code in its box. (6) Enter the Focal Group Code in its box. (7) Enter the number of levels of total scores to analyze in the corresponding box. (8) For each level, enter the minimum and maximum scores. Click the scroll bar to go to the next level. You may need to click the down end of the scroll bar to correct errors or change minimum and maximum values for a level.
|
||||
139=Directions: In polynomial regression smoothing, the value of a point y at a given time t is estimated by the sum of regression weights times t raised to a power of 1, 2, etc. up to the order specified. Enter the order and click the OK button.
|
||||
140=Directions: Cases should consist of k dichotomous item scores (0 and 1 scores.) You can use the Classical Test program to score your test and save the item scores to the grid if necessary.
|
||||
141=Directions: First, click on the variable name that represents the sample lot number. Next, click on the variable that represents the measurement. Click on the sigma button to change the default and click on any of the optional check boxes and enter specifications desired. Click the Compute button to obtain the results. Up to 200 groups may be analyzed. Note! Equal group sizes of 2 to 25 required for ranges analysis. Control limits are plus and minus 3 sigma.
|
||||
142=R = 1 - (s2 / S2) x (1 - r) where R is the estimated reliability of a test obtained on a new group with variance S2 when a reliability of r was obtained for the same test on a group with variance s2. It is assumed the difference in variance is due soley to the difference in true score variance of the two groups. See Theory of Mental Tests by H. Guliksen, 1950.
|
||||
143=Directions: Your data grid should consist of a table of N rows and M+1 column variables. Each row should have a string type label variable and M columns of integer frequency data. 1. Enter the variable for the row labels (strings) 2. Enter the variables representing the columns of frequency integers 3. Select the Options desired 4. If only one variable is to be considered the reference variable, click the button labeled "Use Only the reference variable selected and click on one of the column variables just selected to represent the reference distribution. 5. If each column variable is to be considered as a reference variable, click on the other button labeled "Let each variable be a reference variable" 6. Change the alpha level for significance if desired. 7. Check the Bonferroni contrasts if desired. 8. Click the Compute button to obtain the results.
|
||||
144=This procedure calculates the Pearson Product-Moment correlation coefficients for two or more variables. If one or more of the variables selected have been filtered out or contain a missing value, the case containing that variable will not be included in the analysis (list-wise deletion.) You may elect to obtain not only the correlations but also the raw cross-products, the variance-covariance matrix and the means, variances and standard deviations of the variables. Click on the variable in the list to the left and enter it for analysis by clicking the right arrow box. Repeat this for each variable to be included or click the ALL button to include all variables.
|
||||
145=Directions: First click on the variable name that represents the sample lot number. Next, click on the variable that represents the measurement. Click the Compute Button to obtain the results. NOTE! Equal group (lot) sizes of 2 to 25 required for Sigma analysis. Control limits are plus and minus 3 sigma. Up to 200 lots may be analyzed.
|
||||
146=Description: Straight Line Depreciation calculates the depreciation allowance for an asset over one period in it's life. The function divides the cost minus the salvage value by the number of years of useful life of the asset. Cost is the inital amount paid for the asset. Salvage is the value left at the end of the asset's life. EXAMPLE: What is the depreciation value on might expect for a computer purchased for $2,000.00 and expected to have a useful life of three years with no residual value? ANSWER: Approximately $666.67
|
||||
147=R = Kr / (1 + (K - 1) r where R is the estimated reliability of a test when increased by a factor of K. K is the number of items in the lengthened test divided by the number of items in the original test. r is the reliability of the original test.
|
||||
148=Directions: Click on the variables from the left list of available variables. Click the right-pointing arrow to enter your selection(s). You can remove a selected variable by clicking on it and click the left-pointing arrow button. Click the Compute button to do the analysis. NOTE: Some leaves may represent fragments smaller than the leaf depth.
|
||||
149=Description: Sum of Years Digits Depreciation calculates depreciation amounts for an asset using an accelerated depreciation method. This allows for higher depreciation in the early years of an asset's life. Cost is the initial cost of the asset. Salvage is the value of the asset at the end of it's life expectancy. Life is the length of the asset's life expectancy. Period is the period that you wish to calculate the depreciation. EXAMPLE: What is the depreciation for period 1, 2 or 3 that one can claim for a computer purchased at a price of $2,000.00 and expected to have a useful life of 3 years with no salvage value? ANSWER: $1,000.00 the first year, $666.67 the second period and $333.33 the last year.
|
||||
150=New variables may be created that are transformations of an existing variable or a combination of two variables or a variable and a constant. For example, you may want to create a new variable that is the natural log of an existing variable. As another example, you may want to create a variable that is the product of two other variables. To create the new variable, enter a name for the new variable in the edit box provided for the new variable name. Next, select the transformation in the list of functions available. The selected transformation will be shown in a box below the list of functions. Next, click on the name of the variable for the first arguement of the function to be performed and use the corresponding right arrow button to enter it. If a second variable is required (V2) click on the name of the variable and enter it with the corresponding arrow for V2. If a constant is required, click on the constant edit box and enter the value. Click on the Compute button.
|
||||
151=Directions: For Dependent samples, click on the three variables representing X, Y and Z (in that order.) The test will compare the r(x,y) with the r(x,z). For Independent samples, click on the X and Y variables to be correlated and then the variable representing the group coding variable. The correlations obtained in each of two groups will be compared.
|
||||
152=Directions: For independent groups you should have a variable indicating group membership using 1 and 2 for the group codes and a variable with 0 or 1 values which represent observed or not observed in the group. For dependent proportions you should have two variables code with 0 or 1 in each case.
|
||||
153=Select the Dependent Variable and enter it in its box. Select the predictors (including the ones dependent on the instrumental variables) and enter them in the explanatory list. Copy the predictors dependent on the instrumental variables to the Instrumental Variables list. Add the instrumental variables to the same list. Select options desired and click the Compute button. NOTE: The number of variables in the Instrumental list should be equal to or greater than the Explanatory list.
|
||||
154=Directions: Data may be entered on this form or from a file loaded in the grid. First, enter the number of rows and columns pressing the return key after each entry. If entering Grid data, click on the variables corresponding to row, column and frequency data. If entering on this form, enter the frequencies in the cells corresponding to the row and column of your data.
|
||||
155=Directions: Click on the variable that represents the count of defects. Enter the number inspected in each subgroup (lot.) Note - all groups are of equal size. Click on a Sigma button to change to a different value. You may enter a specific value if you choose the X sigma option. Click the Compute button to see the results.
|
||||
156=Directions: First click on one of the variables representing matched pairs of observations from the list of available variables. Click the right-pointing button to enter your choice for variable 1. Repeat for the second variable. Click the Compute button to obtain the results.
|
||||
157=Directions: The repeated measures ANOVA requires you to select two or more variables (columns) which represent repeated observations on the same subjects (rows.) Homogeneity of variance and covariance are assumed and may be tested as an option. In addition, the ANOVA provides the basis for estimates of reliability as developed by Hoyt (Intraclass reliability) with the adjusted estimate equivalent to the Cronbach Alpha estimate. Finally, you may elect to plot the means obtained for the repeated measures.
|
||||
158=Weighted Least Squares Regression lets you save the residuals and squared residuals for an OLS weighted analysis. You may also complete a regression of these residuals on the independent variables and save the residuals and squared residuals from those analyses. The square root of the reciprocal of the absolute squared residuals from this last analysis may be used as weights to reduce the heteroscedasticity in your data. If this option is chosen, an OLS regression of the weighted variables is conducted. This may be done through the origin.
|
||||
159=Directions: Firs, click in the variable name that represents the sample lot numbers. Next, click on the variable that represents the measurement. Click on the sigma button to change the default and click on any of the optional check boxes and enter specifications desired. Click the Compute button to obtain the results.
|
||||
160=Correspondence analysis is a method for examining the relationship between two sets of categorical variables much as in a Chi-Squared analysis of a two-way contingency table. In fact, a typical chi-squared analysis is completed as part of this procedure. In addition, visualization of the relationships among the columns or rows of the analysis is performed in a manner similar to factor analysis. The data analyzed in the visualization is the table of relative proportions, that is, the original frequency values divided by the sum of all frequencies. The relative proportions of the row sums and the column sums are termed the �masses� of the rows or columns. The method used to analyze the relative proportions involves what is now called the �Generalized Singular Value Decomposition� or more simply the generalized SVD. This method obtains roots and vectors of a rectangular matrix by decomposing that matrix into three portions: a matrix of left singular column vectors (A) that has n rows and q columns (n � q), a square diagonal matrix with q rows and columns of singular values (D), and a transposed matrix (B�) that is m x q in size of right generalized singular vectors (m = q-1). Completing this analysis involves several steps. The first is to obtain the (regular) SVD analysis of a matrix Q defined as Dr-1/2PDc -1/2 where Dr and Dc are diagonal matrices of row and column relative proportions and P is the matrix of relative proportions. The SVD of Q gives Q = U D V� where D is the desired diagonal matrix of eigenvalues and U�U = V�V = I. It should be noted that the first of the q roots is trivial and to be ignored. At this point we obtain A = Dr1/2U and B = Dc 1/2 V. The results of this SVD analysis is available on the output. Now P = ADB�. The row coordinates F and column coordinates G are then computed according to the table: Analysis Choice Button Selected Row Coordinates Column Coordinates Row Profile Row F = Dr-1AD G = Dc-1B Column Profile Column F = Dr-1A G = Dc-1BD Both Profiles Both F = Dr-1AD G = Dc-1BD If Row profiles are computed, the row coordinates are weighted centroids of the column coordinates and the inertias D2 refer only to the row points. If the column profiles are computed, the column coordinates are weighted eentroids of the row coordinates and the inertias D2 refer only to the column points. If both profiles are selected, neither row or column coordinates are weighted centroids of the other but the inertias D2 refer to both sets of points. The q-1 inertias are plotted in a manner similar to a scree plot of roots in a factor analysis. The total inertia is, in fact, the chi-squared statistic divided by the total of all cell frequencies. You may elect to plot the coordinates for any two pairs of coordinates. This will provide a graphical representation of the separation of the row or column categories similar to a plot of variables in a discriminant function analysis or factors in a factor analysis. A way of looking at correspondence analysis is to consider it as a method for decomposing the overall inertia by identifying a small number of dimensions in which the deviations from the expected values can be represented. This is similar to factor analysis where the total variance is decomposed so as to arrive at a lower dimensional representation of variables.
|
||||
@ -697,6 +697,7 @@ begin
|
||||
AMsg := 'Numeric input required for alpha level.';
|
||||
exit;
|
||||
end;
|
||||
|
||||
if (tmp <= 0) or (tmp >= 1) then
|
||||
begin
|
||||
AControl := AlphaEdit;
|
||||
|
||||
Reference in New Issue
Block a user