Statistics and data analysis

1. What is meant by confidence interval in data analysis e.g. 95% confidence interval?
2. How does p<0.05 estimate significant difference?

David Quinn answered for the first point: the data have a 95% probability to be in the interval.

I am not sure to understand the second question. I assume that you speak of the p-value. The p-value is the result of an hypothesis test: it is the probability that the data are what they are if the null hypothesis is true.

Imagine that you have two sets of numbers, generated by unknown random processes, and you want to know if the difference between the averages of the two sets is statistically significant.

You perform an hypothesis test where the null hypothesis is that both random processes have the same mean. If your p-value is less than 5%, it means that given your two sets, there is less than 5% probability that both random processes have the same mean. So you can say that the difference of average between the two sets of numbers is statistically significant.

1. An interval, based on a sample, which has a $95$% probability of containing the unknown population parameter. Note that the interval is variable, but the unknown parameter is fixed.
2. Not sure what you mean. the word "significant" is used to describe the p value of 5%, just like "highly significant" is associated with a p value of 1%.
• Ok, thanks for your answer. First of all I wish to add that I am an amateur in statistics, I'm just trying to find my way around data analysis as it's caught my interest... What do you mean by 'unknown population parameter' and 'interval' when you say the interval is variable. 2.) would you explain what is meant by 'P value'? – user255723 Jul 20 '15 at 16:43
• I'm not sure if this is the right place for a full explanation, and there are numerous sources online or in textbooks you can consult. Examples of an unknown population parameter: the mean of a normal distribution or the proportion of a binomial distribution which we want to estimate. The interval is variable because it depends on what sample we obtain to make the confidence interval. I hope this helps. – David Quinn Jul 20 '15 at 16:56
• Ok, where would be more convenient, I like wechat, do you mind conversing there? Meanwhile I would appreciate recommendations to other sources. – user255723 Jul 20 '15 at 17:08
• Sure we can wechat – David Quinn Jul 20 '15 at 17:45
• Ok send me your linked in contact I'll send my wechat id there – user255723 Jul 20 '15 at 18:23