Hi, I have a sample data set as follows:
I want to calculate c-alpha (Cronbach's alpha) for the items from a1 through a21. However, there are missing values and "0"s. I may not impute the missing values. They have to remain as missing. So, I am not sure to use "casewise" option. I do not have any idea when I don't use this option whether the calculation of Stata is theoretically true.
Moreover, I did not understand how Stata calculates the scale (gen(sc)) when I use "casewise" option or don't use it, for each observation (row). I read the manual entry bu it is still unclear for me. Thank you in advance for your answers.
Demet
Code:
* Example generated by -dataex-. To install: ssc install dataex clear input float(a1 a2 a3 a4 a5 a6 a7 a8 a9 a10 a11 a12 a13 a14 a15 a16 a17 a18 a19 a20 a21) .9527102 .9957728 .9974957 . . . .5970149 . . . . .7324561 .4298246 .9800444 .0712788 .9466019 .0552764 . . . . .9402389 .083014 .3215097 0 .5517241 .9746193 .9452736 .9548023 .7268518 .2929093 .4291667 .3114035 .2587719 .2084257 .067086 .4611651 .2914573 .1024096 .9532164 .8878924 .9910314 .7625723 .99694 .9843503 .7415385 .8793104 .8426396 . .9039548 .7175926 .9599829 .8291667 .1184211 .2631579 .7760532 .8721174 .1067961 .7688442 .6144578 .251462 .9775785 .0986547 .1455423 .7412004 .8068792 . . .4467005 .3034826 . .675926 .3648897 .4875 .4298246 .5219298 .7006652 .9098532 .7766991 .1557789 . . .206278 .3273543 .7406301 .0247582 .050907 .7046154 .8275862 .4111675 .5472637 .4124294 .2361111 .9748674 .6666666 .1578947 .0877193 .4345898 .721174 .2475728 .4120603 .7831326 .4385965 .5156951 .6950673 .8660333 .9995263 .9931427 .8584616 .9770115 .715736 .4029851 .7457627 .337963 .9876447 .6875 .8508772 .7061403 .7804878 .7044025 .4417475 .0251256 .4939759 .4444444 .8699552 .2107623 .1280848 .0185151 .0263069 . .1206897 .8832487 .4328358 .9152542 .8287037 .5884128 .7291666 .9298246 .9429824 .1552106 .3563941 .0291262 .201005 0 .0116959 .4798206 .2242152 .0358467 .0183637 .0246961 .3969231 .6091954 .8629442 .8507463 .8022599 .712963 .9056951 .3083333 .6973684 .7719298 .5942351 .7253669 .6796117 .2512563 .2771084 .4502924 .632287 .6233184 .3831971 .0262813 .0591545 0 .1494253 .9441624 .8706468 .8757062 .5185185 .9036292 .2166666 .0570175 .0263158 .172949 .9601677 .9708738 .2864321 .1325301 .871345 .4529148 .9013453 .0000184 .2792412 .4849365 . .8390805 .071066 .1791045 .079096 .4212963 .9849488 .8041667 .8289474 .8201754 .6208426 .6666666 .9029126 .160804 .8072289 .6666667 .5919282 .4798206 .9175144 .058476 .1720804 .3723077 .5689655 .8883249 .9552239 .8813559 .4305556 .9767663 .5916667 .6710526 .6885965 .7627494 .9559748 .1165048 .7336683 .6686747 .3508772 .9058296 .367713 end
I want to calculate c-alpha (Cronbach's alpha) for the items from a1 through a21. However, there are missing values and "0"s. I may not impute the missing values. They have to remain as missing. So, I am not sure to use "casewise" option. I do not have any idea when I don't use this option whether the calculation of Stata is theoretically true.
Moreover, I did not understand how Stata calculates the scale (gen(sc)) when I use "casewise" option or don't use it, for each observation (row). I read the manual entry bu it is still unclear for me. Thank you in advance for your answers.
Demet
