Вы находитесь на странице: 1из 2

Cp and Pp.

Cp indices recognize the fact that your samples represent rational subgroups,
which indicate how the process would perform if the shift and drift between
subgroups could be eliminated. Therefore, it calculates process spread using
within-subgroup variation.

Pp, on the other hand, ignores subgroups and considers the overall variation
of the entire process. This overall variation accounts for the shift and drift
that can occur between subgroups; therefore, it is useful in measuring
capability over time. If your Pp value differs greatly from your Cp value, you
conclude that there is significant variation from one subgroup to another.

Compare your Cp and Pp values to benchmark values to determine whether


to improve your process; many industries use benchmark values of 1.33. If
the values of your indices are less than the benchmark values, you must take
steps to improve your process.

For example, a butter manufacturer uses an automatic filling machine to


make 500 gm containers of butter. The manager wants to ensure that the
filling process meets specifications of 498 gm to 502 gm of butter in each
container. To assess the capability of this process, the manager randomly
collects and weighs five containers per day for twenty days: each day's
collection is a subgroup for this analysis. Further inspection reveals a leak in
the machine, so the manager fixes this problem and collects another twenty
subgroups of 5 containers after the improvement takes effect. The following
graphs illustrate Cp and Pp values of this filling process before and after
improvement.

P value:

Determines the appropriateness of rejecting the null hypothesis in a


hypothesis test. P-values range from 0 to 1. The smaller the p-value, the
smaller the probability that rejecting the null hypothesis is a mistake. Before
conducting any analyses, determine your alpha (a) level. A commonly used
value is 0.05. If the p-value of a test statistic is less than your alpha, you
reject the null hypothesis.

Because of their indispensable role in hypothesis testing, p-values are used in


many areas of statistics including basic statistics, linear models, reliability,
and multivariate analysis among many others. The key is to understand what
the null and alternate hypotheses represent in each test and then use the p-
value to aid in your decision to reject the null.

For example, consider a 2-sample t-test where you are testing the difference
between the mean strength of steel from two mills based on random samples
from each. In this case, the null hypothesis states that the two population
means are equal while the alternate hypothesis states that they are not equal.
A p-value below your cutoff level suggests that the population means are
different.

Suppose you are also conducting regression analyses on steel strength where
temperature is one of the explanatory variables. You will see a p-value for
each regression coefficient. Here, the default test is to determine if the
estimated coefficient for temperature is different from zero. Therefore, the
null hypothesis states that the coefficient equals zero while the alternate
hypothesis states that it is not equal to zero. A p-value below your cutoff
level suggests that the coefficient for temperature is significantly different
from zero and likely to be a meaningful addition to your model.

The p-value is calculated from the observed sample and represents the
probability of incorrectly rejecting the null hypothesis when it is actually true
(Type I error). In other words, it is the probability of obtaining a difference at
least as large as the one between the observed value and the hypothesized
value through random error alone.

Вам также может понравиться