Summary:

Accounts for extraneous variables (or alternative explanations) of the independent variable on the dependent variable

History Effect

(Threat to Internal Validity)

Summary:

Events that occur between pretest and posttest of a study that could affect participants in ways that impact the dependent variable

(i.e. stress, natural disasters, news events, illness)

Solution:

Use a control group and experiment group to compare for differences

(people not exposed to event(s))

Maturation

(Threat to Internal Validity)

Summary:

Changes noticed in participants because of elapsed time since beginning of study

(i.e. older, wiser, stronger, experienced)

(Usually occurs with studies of children)

Solution:

Use control group and experiment group to compare for differences

Testing

(Threat to Internal Validity)

Summary:

Participants may become "test-wise" (doing better) on posttest from a pretest in attempt to obtain baseline

Solution:

Use control group that does NOT receive pretesting

Instrumentation

(Threat to Internal Validity)

Summary:

Occurs if measurement tools are NOT accurate or procedures are NOT standardized

Solution:

          1. Physical instruments must be calibrated to give   consistent and accurate results

 2. Individuals must be well trained and follow consistent procedures

          3. Written instruments must be carefully designed to ask correct questions

Selection Bias

(Threat to Internal Validity)

Summary:

Participants selected in a NONRANDOM manner could introduce possibilities that experimental and control groups consist of differences before study begins that excludes any effects of the independent variable

Solution:

1. Randomly assign recruited volunteers to groups

  2. Randomly assign participants to groups based on matching participants on selected characteristics

 3. Pretesting groups on measures of the dependent variable to make sure there are not any pre intervention differences between groups

Selection Maturation Effect

(Threat to Internal Validity)

Summary:

Occurs when using intact groups that vary in maturation levels

(Combination of Selection Bias & Maturation Effects)

Solution:

Pretest and\or prescreen groups on maturity levels

Statistical Regression

(Threat to Internal Validity)

Summary:

Occurs when participants are selected based on high or low scores

Solution:

Select random sample of participants representing full range of scores rather than selectinng participants based on high and low scores would account for regression of the mean

Mortality\Attrition

(Threat to Internal Validity)

Summary:

(Mortality Effect)

Loss of participants causes problems with group comparisons when number of participants in some groups falls below established guidelines for statistical analysis

(i.e. death, poor health, location changes, dropouts)

Solution:

1. Oversampling & use large group sizes

2. Offer incentives to encourage participants to stay in the study

3. Obtain good demographic info of participants at the beginning of study and determine if group makeup changed at the end of study after mortality effect

Hawthorne Effect

(Threat to Internal Validity)

Summary:

Occurs when participants' attitudes toward being observed or involved in a study affect their behavior

Solution:

1. Provide control grroup with "special attention" comparable to the experimental group, but make sure it does NOT directly impact the dependent variable

2. Keep participants from knowing they are taking part in a study

(Both are difficult solutions)

Placebo Effect

(Threat to Internal Validity)

Summary:

Regarding medication studies, this effect is caused by participants' expectations of intervention(s) rather than any actual provided intervention(s), control groups are given an inactive substance a.k.a. placebo

(Blind Study)

Participants are NOT aware whether or not they are receiving placebo or actual intervention

(Double Blind)

Participants and researcher(s) are NOT aware of who receives placebo or actual intervention

Solution:

1. Both control & experimental groups receive same info about study to elicit similar expectations

2. Both control & experimental groups receive as little info as possible about the study to reduce expectations

Diffusion of Treatment

(Threat to Internal Validity)

Summary:

Occurs when the intervention applied to one group spills over or contaminates another group

Solution:

1. Attempt to use same population with matching demographics in different locations

2. In case solution one cannot be done, explain study to participants emphasizing to keep intervention and study info confidential until end of study

Location

(Threat to Internal Validity)

Summary:

Occurs when there are differences in the locations where interventions take place

(i.e. room lighting, wall decor, windows or lack of)

Solution:

1. Make locations the same for all participants

2. If solution one is NOT feasible, attempt to minimize location differences

Implementation

(Threat to Internal Validity)

Summary:

Possibility individual(s) responsible for implementing experimental intervention may inadvertently introduced inequality or bias into the study

Solution:

1. All persons responsible for implementation should be equally trained, competent, and follow protocol

2. All persons responsible for implementation should practice and instruct use of measurement tool

3. Avoid favorite bias by assigning tasks to neutral presenters\observers

Selection Treatment Interaction

(Threat to External Validity)

Summary:

Concerns ability of a researcher to generalize the results of a study beyond the groups involved in the study

(i.e. When using intact groups, researchers have no way of knowing if results are truly representative of a larger population)

Solution:

Results can legitimately be generalized to the study population but NOT the population when a study uses a random sample of participants

Setting Treatment Interaction

(Threat to External Validity)

Summary:

The extent to which environmental conditions or setting under which a study was conducted can be duplicated in other settings

Solution:

Attempt to account for the threat of internal validity in relation to the location effect to minimize differences in both the participants (knowledge, socioeconomic status, support networks, etc.) and settings (cultural climate, privacy, previous issues with what the study is on, etc.)

History Treatment Interaction

(Threat to External Validity)

Summary:

Occurs when researchers try to generalize findings to past and future situations

Solution:

Attempt to account for threats of internal validity, but specifically the history effect as well as location effect